diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,90260 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999806054964023,
+  "eval_steps": 500,
+  "global_step": 12890,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.519832075963106,
+      "learning_rate": 5.16795865633075e-08,
+      "loss": 1.0553,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.888495993953021,
+      "learning_rate": 1.03359173126615e-07,
+      "loss": 1.0691,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.777014625287366,
+      "learning_rate": 1.5503875968992249e-07,
+      "loss": 1.0115,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.288166885816697,
+      "learning_rate": 2.0671834625323e-07,
+      "loss": 1.0531,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.728059218918297,
+      "learning_rate": 2.583979328165375e-07,
+      "loss": 1.0227,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.903468059340747,
+      "learning_rate": 3.1007751937984497e-07,
+      "loss": 1.071,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.673274081631383,
+      "learning_rate": 3.6175710594315246e-07,
+      "loss": 1.1014,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.127045187299337,
+      "learning_rate": 4.1343669250646e-07,
+      "loss": 1.0198,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.42937660527831,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 1.0276,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.625662352763816,
+      "learning_rate": 5.16795865633075e-07,
+      "loss": 0.9806,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.336066643354788,
+      "learning_rate": 5.684754521963825e-07,
+      "loss": 1.0316,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.462067474116861,
+      "learning_rate": 6.201550387596899e-07,
+      "loss": 1.0035,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.403141428987006,
+      "learning_rate": 6.718346253229975e-07,
+      "loss": 1.0206,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.128595347934199,
+      "learning_rate": 7.235142118863049e-07,
+      "loss": 0.9459,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.395879777195401,
+      "learning_rate": 7.751937984496125e-07,
+      "loss": 0.9791,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.316994153124984,
+      "learning_rate": 8.2687338501292e-07,
+      "loss": 0.8712,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.630215579253789,
+      "learning_rate": 8.785529715762274e-07,
+      "loss": 0.9682,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.294689706674586,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 0.9027,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.422215964176046,
+      "learning_rate": 9.819121447028424e-07,
+      "loss": 0.9334,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3414770939835425,
+      "learning_rate": 1.03359173126615e-06,
+      "loss": 0.9431,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.195632731985836,
+      "learning_rate": 1.0852713178294575e-06,
+      "loss": 0.934,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.612591981865614,
+      "learning_rate": 1.136950904392765e-06,
+      "loss": 0.9314,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6314997357443173,
+      "learning_rate": 1.1886304909560723e-06,
+      "loss": 0.8186,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.764521836778098,
+      "learning_rate": 1.2403100775193799e-06,
+      "loss": 0.8507,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9395074680919357,
+      "learning_rate": 1.2919896640826874e-06,
+      "loss": 0.8617,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.609294136301998,
+      "learning_rate": 1.343669250645995e-06,
+      "loss": 0.8471,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4551541628476996,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 0.8961,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.71237206223252,
+      "learning_rate": 1.4470284237726098e-06,
+      "loss": 0.8651,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.414596350445245,
+      "learning_rate": 1.4987080103359176e-06,
+      "loss": 0.8452,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.522648076681847,
+      "learning_rate": 1.550387596899225e-06,
+      "loss": 0.8308,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.482017790337925,
+      "learning_rate": 1.6020671834625322e-06,
+      "loss": 0.7951,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.694296117241535,
+      "learning_rate": 1.65374677002584e-06,
+      "loss": 0.8104,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2735440468482206,
+      "learning_rate": 1.7054263565891473e-06,
+      "loss": 0.878,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.460717666737549,
+      "learning_rate": 1.7571059431524549e-06,
+      "loss": 0.7748,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.615230363724519,
+      "learning_rate": 1.8087855297157624e-06,
+      "loss": 0.8503,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0739543624168046,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 0.7561,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.098811848895224,
+      "learning_rate": 1.9121447028423773e-06,
+      "loss": 0.7309,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4512717663597585,
+      "learning_rate": 1.963824289405685e-06,
+      "loss": 0.8263,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2516002262721893,
+      "learning_rate": 2.0155038759689923e-06,
+      "loss": 0.8068,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.596235532100443,
+      "learning_rate": 2.0671834625323e-06,
+      "loss": 0.8449,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.545090219663501,
+      "learning_rate": 2.1188630490956074e-06,
+      "loss": 0.8061,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.363567341283558,
+      "learning_rate": 2.170542635658915e-06,
+      "loss": 0.8206,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.163689095682903,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.7695,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2080585620688122,
+      "learning_rate": 2.27390180878553e-06,
+      "loss": 0.7675,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3873360143137075,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 0.7691,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.494632490226066,
+      "learning_rate": 2.3772609819121447e-06,
+      "loss": 0.7957,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.244123945182519,
+      "learning_rate": 2.4289405684754527e-06,
+      "loss": 0.7594,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.277922201876151,
+      "learning_rate": 2.4806201550387598e-06,
+      "loss": 0.7248,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0354233345018833,
+      "learning_rate": 2.5322997416020673e-06,
+      "loss": 0.7247,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1956830238083183,
+      "learning_rate": 2.583979328165375e-06,
+      "loss": 0.7942,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1086478781492746,
+      "learning_rate": 2.635658914728683e-06,
+      "loss": 0.7626,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.212343665165485,
+      "learning_rate": 2.68733850129199e-06,
+      "loss": 0.7707,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.185894513201737,
+      "learning_rate": 2.7390180878552975e-06,
+      "loss": 0.7752,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.391570359350091,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 0.7444,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.625443466510659,
+      "learning_rate": 2.842377260981912e-06,
+      "loss": 0.7962,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8983609533379908,
+      "learning_rate": 2.8940568475452197e-06,
+      "loss": 0.7051,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1952030249616508,
+      "learning_rate": 2.9457364341085276e-06,
+      "loss": 0.7571,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.077316241532325,
+      "learning_rate": 2.997416020671835e-06,
+      "loss": 0.7287,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.227086591108584,
+      "learning_rate": 3.0490956072351423e-06,
+      "loss": 0.7853,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2554648050155466,
+      "learning_rate": 3.10077519379845e-06,
+      "loss": 0.7785,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1596438471020365,
+      "learning_rate": 3.1524547803617574e-06,
+      "loss": 0.8295,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.282697453560096,
+      "learning_rate": 3.2041343669250645e-06,
+      "loss": 0.7639,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.758532025047504,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 0.687,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2299768870913597,
+      "learning_rate": 3.30749354005168e-06,
+      "loss": 0.7668,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.111093569844413,
+      "learning_rate": 3.3591731266149875e-06,
+      "loss": 0.8231,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1534840631827485,
+      "learning_rate": 3.4108527131782946e-06,
+      "loss": 0.7143,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.216368458223642,
+      "learning_rate": 3.462532299741602e-06,
+      "loss": 0.7461,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3280494216837937,
+      "learning_rate": 3.5142118863049097e-06,
+      "loss": 0.7476,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.241321356552836,
+      "learning_rate": 3.5658914728682177e-06,
+      "loss": 0.7498,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1221608661794136,
+      "learning_rate": 3.617571059431525e-06,
+      "loss": 0.7522,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0474193518921426,
+      "learning_rate": 3.6692506459948323e-06,
+      "loss": 0.8124,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9518444288280057,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 0.7243,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1636466718184812,
+      "learning_rate": 3.772609819121447e-06,
+      "loss": 0.7377,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.856977439681704,
+      "learning_rate": 3.8242894056847545e-06,
+      "loss": 0.7054,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0053756710099635,
+      "learning_rate": 3.875968992248063e-06,
+      "loss": 0.7338,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.158365731996466,
+      "learning_rate": 3.92764857881137e-06,
+      "loss": 0.7236,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4102976060464267,
+      "learning_rate": 3.979328165374677e-06,
+      "loss": 0.7754,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.17610689256887,
+      "learning_rate": 4.031007751937985e-06,
+      "loss": 0.7185,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0942309382339497,
+      "learning_rate": 4.082687338501292e-06,
+      "loss": 0.6833,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2703418958195645,
+      "learning_rate": 4.1343669250646e-06,
+      "loss": 0.7626,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3007004210388113,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 0.7403,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8774740106252445,
+      "learning_rate": 4.237726098191215e-06,
+      "loss": 0.7147,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5136139217593367,
+      "learning_rate": 4.289405684754522e-06,
+      "loss": 0.7195,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1863608527720064,
+      "learning_rate": 4.34108527131783e-06,
+      "loss": 0.7224,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2834091419898668,
+      "learning_rate": 4.3927648578811375e-06,
+      "loss": 0.7566,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.386091234092043,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.7626,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3674971878373423,
+      "learning_rate": 4.4961240310077525e-06,
+      "loss": 0.6586,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.172996505874935,
+      "learning_rate": 4.54780361757106e-06,
+      "loss": 0.6604,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9070537311083688,
+      "learning_rate": 4.599483204134368e-06,
+      "loss": 0.6725,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9524328607599153,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 0.7146,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0347277873038725,
+      "learning_rate": 4.702842377260982e-06,
+      "loss": 0.7531,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.209982622000607,
+      "learning_rate": 4.754521963824289e-06,
+      "loss": 0.7188,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9692459087380176,
+      "learning_rate": 4.806201550387598e-06,
+      "loss": 0.6813,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1049024850181413,
+      "learning_rate": 4.857881136950905e-06,
+      "loss": 0.743,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1798186450732397,
+      "learning_rate": 4.909560723514212e-06,
+      "loss": 0.7513,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1645572248346174,
+      "learning_rate": 4.9612403100775195e-06,
+      "loss": 0.7533,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0228318887762877,
+      "learning_rate": 5.012919896640828e-06,
+      "loss": 0.6592,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.206621682303021,
+      "learning_rate": 5.064599483204135e-06,
+      "loss": 0.7562,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1311466852020433,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.7061,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9809595587086415,
+      "learning_rate": 5.16795865633075e-06,
+      "loss": 0.7214,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9387466972711545,
+      "learning_rate": 5.219638242894057e-06,
+      "loss": 0.7502,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8229280598602298,
+      "learning_rate": 5.271317829457366e-06,
+      "loss": 0.6942,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8532957824955443,
+      "learning_rate": 5.322997416020672e-06,
+      "loss": 0.732,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9006885008912018,
+      "learning_rate": 5.37467700258398e-06,
+      "loss": 0.6513,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9933017382249472,
+      "learning_rate": 5.4263565891472865e-06,
+      "loss": 0.7436,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0787408819589306,
+      "learning_rate": 5.478036175710595e-06,
+      "loss": 0.6851,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9692236315472937,
+      "learning_rate": 5.529715762273902e-06,
+      "loss": 0.7102,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0958519249073144,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 0.6779,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1179506397493193,
+      "learning_rate": 5.6330749354005176e-06,
+      "loss": 0.7571,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9953179391474662,
+      "learning_rate": 5.684754521963824e-06,
+      "loss": 0.6988,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.209611906437212,
+      "learning_rate": 5.736434108527133e-06,
+      "loss": 0.7277,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9401680204051799,
+      "learning_rate": 5.788113695090439e-06,
+      "loss": 0.6803,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1188809424704016,
+      "learning_rate": 5.839793281653747e-06,
+      "loss": 0.7199,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0943100592990525,
+      "learning_rate": 5.891472868217055e-06,
+      "loss": 0.6849,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2713837146185636,
+      "learning_rate": 5.943152454780362e-06,
+      "loss": 0.7577,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.127209132176853,
+      "learning_rate": 5.99483204134367e-06,
+      "loss": 0.7871,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.034930723731037,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 0.7148,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.136418574307478,
+      "learning_rate": 6.0981912144702846e-06,
+      "loss": 0.6921,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.007500412728349,
+      "learning_rate": 6.149870801033592e-06,
+      "loss": 0.6471,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0923841205992715,
+      "learning_rate": 6.2015503875969e-06,
+      "loss": 0.7567,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.266590622953997,
+      "learning_rate": 6.253229974160208e-06,
+      "loss": 0.6969,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.061399223889906,
+      "learning_rate": 6.304909560723515e-06,
+      "loss": 0.7352,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9181609203148584,
+      "learning_rate": 6.356589147286822e-06,
+      "loss": 0.6856,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1843507167792544,
+      "learning_rate": 6.408268733850129e-06,
+      "loss": 0.6914,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.226492897418139,
+      "learning_rate": 6.459948320413437e-06,
+      "loss": 0.7682,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9479308374334845,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 0.6866,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9634312430850824,
+      "learning_rate": 6.563307493540052e-06,
+      "loss": 0.6599,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0373930546127754,
+      "learning_rate": 6.61498708010336e-06,
+      "loss": 0.6946,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9244260526837411,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.7157,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.883453956330838,
+      "learning_rate": 6.718346253229975e-06,
+      "loss": 0.6764,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0045545692999402,
+      "learning_rate": 6.7700258397932826e-06,
+      "loss": 0.6809,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9606349757302202,
+      "learning_rate": 6.821705426356589e-06,
+      "loss": 0.7321,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8672674952014412,
+      "learning_rate": 6.873385012919898e-06,
+      "loss": 0.6396,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8517558322844003,
+      "learning_rate": 6.925064599483204e-06,
+      "loss": 0.6908,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0440099335072635,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 0.7415,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.979158649093718,
+      "learning_rate": 7.028423772609819e-06,
+      "loss": 0.6901,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9189054916948147,
+      "learning_rate": 7.080103359173127e-06,
+      "loss": 0.6229,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8979967649651344,
+      "learning_rate": 7.131782945736435e-06,
+      "loss": 0.6727,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.089207860872488,
+      "learning_rate": 7.183462532299742e-06,
+      "loss": 0.6474,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.05975957356314,
+      "learning_rate": 7.23514211886305e-06,
+      "loss": 0.7032,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1621687001374346,
+      "learning_rate": 7.286821705426357e-06,
+      "loss": 0.7346,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.029956346991461,
+      "learning_rate": 7.338501291989665e-06,
+      "loss": 0.7053,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9328938768505268,
+      "learning_rate": 7.390180878552973e-06,
+      "loss": 0.6723,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4819536225184136,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 0.73,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2634611122156376,
+      "learning_rate": 7.493540051679587e-06,
+      "loss": 0.7043,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.03845568283746,
+      "learning_rate": 7.545219638242894e-06,
+      "loss": 0.7747,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.170724138477728,
+      "learning_rate": 7.596899224806202e-06,
+      "loss": 0.7935,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0504183784244647,
+      "learning_rate": 7.648578811369509e-06,
+      "loss": 0.6595,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8734228494690135,
+      "learning_rate": 7.700258397932817e-06,
+      "loss": 0.7251,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1235225253454257,
+      "learning_rate": 7.751937984496126e-06,
+      "loss": 0.6747,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1341870248884347,
+      "learning_rate": 7.803617571059433e-06,
+      "loss": 0.6682,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1759810022127875,
+      "learning_rate": 7.85529715762274e-06,
+      "loss": 0.6936,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1956170030421633,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 0.7492,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.483555592654495,
+      "learning_rate": 7.958656330749354e-06,
+      "loss": 0.7305,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.116766405634247,
+      "learning_rate": 8.010335917312663e-06,
+      "loss": 0.6669,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9162855045739617,
+      "learning_rate": 8.06201550387597e-06,
+      "loss": 0.7261,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.054963532098712,
+      "learning_rate": 8.113695090439278e-06,
+      "loss": 0.7314,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.028434092710246,
+      "learning_rate": 8.165374677002584e-06,
+      "loss": 0.7294,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.048480270719039,
+      "learning_rate": 8.217054263565893e-06,
+      "loss": 0.7383,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2176711992410225,
+      "learning_rate": 8.2687338501292e-06,
+      "loss": 0.7124,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.069919880039789,
+      "learning_rate": 8.320413436692508e-06,
+      "loss": 0.7016,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.205058554739053,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 0.6351,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1284296744738613,
+      "learning_rate": 8.423772609819121e-06,
+      "loss": 0.6765,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9087997643879109,
+      "learning_rate": 8.47545219638243e-06,
+      "loss": 0.7027,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1452848850815083,
+      "learning_rate": 8.527131782945736e-06,
+      "loss": 0.7407,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9991766307667835,
+      "learning_rate": 8.578811369509045e-06,
+      "loss": 0.6794,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9138380269362316,
+      "learning_rate": 8.630490956072353e-06,
+      "loss": 0.6835,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.003442543319759,
+      "learning_rate": 8.68217054263566e-06,
+      "loss": 0.6293,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.95394526232576,
+      "learning_rate": 8.733850129198968e-06,
+      "loss": 0.6773,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.088166035828683,
+      "learning_rate": 8.785529715762275e-06,
+      "loss": 0.7069,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9835559025534641,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 0.7081,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9845261403594612,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.6452,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1896583808017813,
+      "learning_rate": 8.940568475452197e-06,
+      "loss": 0.6978,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.034374240576255,
+      "learning_rate": 8.992248062015505e-06,
+      "loss": 0.7193,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.254205482228632,
+      "learning_rate": 9.043927648578812e-06,
+      "loss": 0.7259,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9883989999974647,
+      "learning_rate": 9.09560723514212e-06,
+      "loss": 0.6599,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0406451751506136,
+      "learning_rate": 9.147286821705427e-06,
+      "loss": 0.6291,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0446696740745978,
+      "learning_rate": 9.198966408268735e-06,
+      "loss": 0.643,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0482308359887362,
+      "learning_rate": 9.250645994832042e-06,
+      "loss": 0.7337,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9553371871357939,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 0.6478,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.924814844515946,
+      "learning_rate": 9.354005167958657e-06,
+      "loss": 0.6263,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.047266915344053,
+      "learning_rate": 9.405684754521964e-06,
+      "loss": 0.6979,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.842052712981622,
+      "learning_rate": 9.457364341085272e-06,
+      "loss": 0.6163,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2294991384925043,
+      "learning_rate": 9.509043927648579e-06,
+      "loss": 0.6639,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1086576018925034,
+      "learning_rate": 9.560723514211887e-06,
+      "loss": 0.6987,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.413299538742816,
+      "learning_rate": 9.612403100775196e-06,
+      "loss": 0.6821,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2113895021622576,
+      "learning_rate": 9.664082687338502e-06,
+      "loss": 0.6325,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.312092100717908,
+      "learning_rate": 9.71576227390181e-06,
+      "loss": 0.7312,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0110742814407456,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 0.7209,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.003722660213722,
+      "learning_rate": 9.819121447028424e-06,
+      "loss": 0.7655,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.066364251847548,
+      "learning_rate": 9.870801033591732e-06,
+      "loss": 0.6983,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.160208896319642,
+      "learning_rate": 9.922480620155039e-06,
+      "loss": 0.7163,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.969679234720519,
+      "learning_rate": 9.974160206718347e-06,
+      "loss": 0.7713,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7571093021627038,
+      "learning_rate": 1.0025839793281656e-05,
+      "loss": 0.6597,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9672538279531169,
+      "learning_rate": 1.0077519379844963e-05,
+      "loss": 0.7103,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1541642868451882,
+      "learning_rate": 1.012919896640827e-05,
+      "loss": 0.7753,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8365651457895535,
+      "learning_rate": 1.0180878552971578e-05,
+      "loss": 0.7249,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9360304766258072,
+      "learning_rate": 1.0232558139534884e-05,
+      "loss": 0.6853,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.096884473617233,
+      "learning_rate": 1.0284237726098191e-05,
+      "loss": 0.7369,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8273847988781062,
+      "learning_rate": 1.03359173126615e-05,
+      "loss": 0.7152,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8844761759607553,
+      "learning_rate": 1.0387596899224808e-05,
+      "loss": 0.6734,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.914089218264573,
+      "learning_rate": 1.0439276485788114e-05,
+      "loss": 0.7143,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.116071906845655,
+      "learning_rate": 1.0490956072351421e-05,
+      "loss": 0.7001,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0417825964367657,
+      "learning_rate": 1.0542635658914731e-05,
+      "loss": 0.7866,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.044203084483132,
+      "learning_rate": 1.0594315245478038e-05,
+      "loss": 0.6179,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0420830230723035,
+      "learning_rate": 1.0645994832041345e-05,
+      "loss": 0.7124,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1051199487865997,
+      "learning_rate": 1.0697674418604651e-05,
+      "loss": 0.7409,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8155533295501942,
+      "learning_rate": 1.074935400516796e-05,
+      "loss": 0.6941,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9123865380632588,
+      "learning_rate": 1.0801033591731266e-05,
+      "loss": 0.7245,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9133835814631164,
+      "learning_rate": 1.0852713178294573e-05,
+      "loss": 0.6549,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7395522368006386,
+      "learning_rate": 1.0904392764857883e-05,
+      "loss": 0.7039,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8978653033690394,
+      "learning_rate": 1.095607235142119e-05,
+      "loss": 0.6929,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0391466846845905,
+      "learning_rate": 1.1007751937984497e-05,
+      "loss": 0.7141,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7644765714244997,
+      "learning_rate": 1.1059431524547803e-05,
+      "loss": 0.6726,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.834831131693005,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 0.6791,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9981262847543069,
+      "learning_rate": 1.116279069767442e-05,
+      "loss": 0.6801,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8607705913420696,
+      "learning_rate": 1.1214470284237727e-05,
+      "loss": 0.6662,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9942224020634212,
+      "learning_rate": 1.1266149870801035e-05,
+      "loss": 0.7746,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9566979647943363,
+      "learning_rate": 1.1317829457364342e-05,
+      "loss": 0.7071,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0894906289355366,
+      "learning_rate": 1.1369509043927648e-05,
+      "loss": 0.7398,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.902596020274279,
+      "learning_rate": 1.1421188630490959e-05,
+      "loss": 0.6448,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8516721411325656,
+      "learning_rate": 1.1472868217054265e-05,
+      "loss": 0.6533,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9422069717376376,
+      "learning_rate": 1.1524547803617572e-05,
+      "loss": 0.7712,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.029117344982057,
+      "learning_rate": 1.1576227390180879e-05,
+      "loss": 0.6953,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9938398502397154,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 0.6658,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.007778419769434,
+      "learning_rate": 1.1679586563307494e-05,
+      "loss": 0.6895,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8139338248104935,
+      "learning_rate": 1.1731266149870802e-05,
+      "loss": 0.6562,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.80716525321881,
+      "learning_rate": 1.178294573643411e-05,
+      "loss": 0.7093,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1476574400631026,
+      "learning_rate": 1.1834625322997417e-05,
+      "loss": 0.7744,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0661453630430664,
+      "learning_rate": 1.1886304909560724e-05,
+      "loss": 0.7099,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1125412525008986,
+      "learning_rate": 1.193798449612403e-05,
+      "loss": 0.7794,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8573275922237884,
+      "learning_rate": 1.198966408268734e-05,
+      "loss": 0.6658,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8576416784089278,
+      "learning_rate": 1.2041343669250647e-05,
+      "loss": 0.6768,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.922463427767487,
+      "learning_rate": 1.2093023255813954e-05,
+      "loss": 0.7218,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9144802338323745,
+      "learning_rate": 1.2144702842377262e-05,
+      "loss": 0.7292,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8917235028415234,
+      "learning_rate": 1.2196382428940569e-05,
+      "loss": 0.6994,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7192218479626784,
+      "learning_rate": 1.2248062015503876e-05,
+      "loss": 0.6272,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.059429194615707,
+      "learning_rate": 1.2299741602067184e-05,
+      "loss": 0.6372,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8141280130390114,
+      "learning_rate": 1.2351421188630493e-05,
+      "loss": 0.7174,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8352882698287405,
+      "learning_rate": 1.24031007751938e-05,
+      "loss": 0.6856,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8149697807255358,
+      "learning_rate": 1.2454780361757106e-05,
+      "loss": 0.6677,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.960900294950215,
+      "learning_rate": 1.2506459948320416e-05,
+      "loss": 0.7238,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.898006562980343,
+      "learning_rate": 1.2558139534883723e-05,
+      "loss": 0.7101,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8680187593575954,
+      "learning_rate": 1.260981912144703e-05,
+      "loss": 0.744,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9090085316303016,
+      "learning_rate": 1.2661498708010338e-05,
+      "loss": 0.6506,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.595526658646269,
+      "learning_rate": 1.2713178294573645e-05,
+      "loss": 0.6481,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6954086909274888,
+      "learning_rate": 1.2764857881136951e-05,
+      "loss": 0.6882,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0163576435791515,
+      "learning_rate": 1.2816537467700258e-05,
+      "loss": 0.7297,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.895852106301838,
+      "learning_rate": 1.2868217054263568e-05,
+      "loss": 0.7747,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9365275003503144,
+      "learning_rate": 1.2919896640826875e-05,
+      "loss": 0.7125,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.708399523251377,
+      "learning_rate": 1.2971576227390181e-05,
+      "loss": 0.6563,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7895994088720608,
+      "learning_rate": 1.302325581395349e-05,
+      "loss": 0.6928,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8279126687923155,
+      "learning_rate": 1.3074935400516796e-05,
+      "loss": 0.7071,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9455123859953396,
+      "learning_rate": 1.3126614987080105e-05,
+      "loss": 0.7194,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8876212984229028,
+      "learning_rate": 1.3178294573643412e-05,
+      "loss": 0.7188,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7747709663808793,
+      "learning_rate": 1.322997416020672e-05,
+      "loss": 0.6819,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.821885113172778,
+      "learning_rate": 1.3281653746770027e-05,
+      "loss": 0.694,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.91953029112429,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7012,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8682867030853527,
+      "learning_rate": 1.3385012919896643e-05,
+      "loss": 0.7418,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7094975512995536,
+      "learning_rate": 1.343669250645995e-05,
+      "loss": 0.669,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9776725818158296,
+      "learning_rate": 1.3488372093023257e-05,
+      "loss": 0.702,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7692678899762055,
+      "learning_rate": 1.3540051679586565e-05,
+      "loss": 0.684,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8705074530572576,
+      "learning_rate": 1.3591731266149872e-05,
+      "loss": 0.6216,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.034166560357307,
+      "learning_rate": 1.3643410852713179e-05,
+      "loss": 0.7239,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9058190781105089,
+      "learning_rate": 1.3695090439276487e-05,
+      "loss": 0.7189,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7473392695345875,
+      "learning_rate": 1.3746770025839795e-05,
+      "loss": 0.6897,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8400541861475124,
+      "learning_rate": 1.3798449612403102e-05,
+      "loss": 0.6924,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6957248590566456,
+      "learning_rate": 1.3850129198966409e-05,
+      "loss": 0.6723,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8766758740788159,
+      "learning_rate": 1.3901808785529717e-05,
+      "loss": 0.7572,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1911608988724023,
+      "learning_rate": 1.3953488372093025e-05,
+      "loss": 0.8032,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7159125748885473,
+      "learning_rate": 1.4005167958656332e-05,
+      "loss": 0.6395,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7821755263454584,
+      "learning_rate": 1.4056847545219639e-05,
+      "loss": 0.6888,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8221975888290423,
+      "learning_rate": 1.4108527131782947e-05,
+      "loss": 0.727,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7422530499772497,
+      "learning_rate": 1.4160206718346254e-05,
+      "loss": 0.6506,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7871411930441092,
+      "learning_rate": 1.421188630490956e-05,
+      "loss": 0.731,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7210350106250718,
+      "learning_rate": 1.426356589147287e-05,
+      "loss": 0.7012,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7810746491438123,
+      "learning_rate": 1.4315245478036177e-05,
+      "loss": 0.6611,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8839496644943539,
+      "learning_rate": 1.4366925064599484e-05,
+      "loss": 0.6647,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8778032963553069,
+      "learning_rate": 1.441860465116279e-05,
+      "loss": 0.6531,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8287017434168242,
+      "learning_rate": 1.44702842377261e-05,
+      "loss": 0.7095,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8776542949642527,
+      "learning_rate": 1.4521963824289408e-05,
+      "loss": 0.7421,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9099247641999344,
+      "learning_rate": 1.4573643410852714e-05,
+      "loss": 0.6702,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8216094934275595,
+      "learning_rate": 1.4625322997416023e-05,
+      "loss": 0.6371,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8987344266291748,
+      "learning_rate": 1.467700258397933e-05,
+      "loss": 0.6714,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7156933741329128,
+      "learning_rate": 1.4728682170542636e-05,
+      "loss": 0.7229,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8087905546016818,
+      "learning_rate": 1.4780361757105946e-05,
+      "loss": 0.7129,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9409079574230799,
+      "learning_rate": 1.4832041343669253e-05,
+      "loss": 0.7056,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6063497311990749,
+      "learning_rate": 1.488372093023256e-05,
+      "loss": 0.6908,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9730694328654426,
+      "learning_rate": 1.4935400516795866e-05,
+      "loss": 0.6481,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5952757469673713,
+      "learning_rate": 1.4987080103359175e-05,
+      "loss": 0.6512,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.98427725723922,
+      "learning_rate": 1.5038759689922481e-05,
+      "loss": 0.7126,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9416110815297036,
+      "learning_rate": 1.5090439276485788e-05,
+      "loss": 0.7322,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8932289917607776,
+      "learning_rate": 1.5142118863049098e-05,
+      "loss": 0.6591,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7060463724441401,
+      "learning_rate": 1.5193798449612405e-05,
+      "loss": 0.6618,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.796732158788336,
+      "learning_rate": 1.5245478036175711e-05,
+      "loss": 0.6555,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6835650023375834,
+      "learning_rate": 1.5297157622739018e-05,
+      "loss": 0.6566,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1006441626998402,
+      "learning_rate": 1.5348837209302328e-05,
+      "loss": 0.7364,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8399146971417395,
+      "learning_rate": 1.5400516795865635e-05,
+      "loss": 0.7592,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.972446906435267,
+      "learning_rate": 1.545219638242894e-05,
+      "loss": 0.7628,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.082671161958546,
+      "learning_rate": 1.550387596899225e-05,
+      "loss": 0.7714,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8677547981650926,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 0.754,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9676800802722405,
+      "learning_rate": 1.5607235142118865e-05,
+      "loss": 0.6773,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9592992010467447,
+      "learning_rate": 1.5658914728682172e-05,
+      "loss": 0.6701,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7348913163105608,
+      "learning_rate": 1.571059431524548e-05,
+      "loss": 0.6879,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.609107874950825,
+      "learning_rate": 1.5762273901808785e-05,
+      "loss": 0.6654,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6397025285151703,
+      "learning_rate": 1.5813953488372095e-05,
+      "loss": 0.6972,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7851211911688332,
+      "learning_rate": 1.5865633074935402e-05,
+      "loss": 0.6876,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9905356346683476,
+      "learning_rate": 1.591731266149871e-05,
+      "loss": 0.7228,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8863281098328457,
+      "learning_rate": 1.5968992248062015e-05,
+      "loss": 0.7152,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.690116408136723,
+      "learning_rate": 1.6020671834625325e-05,
+      "loss": 0.7461,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6336463368807541,
+      "learning_rate": 1.6072351421188632e-05,
+      "loss": 0.6484,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7284714170503186,
+      "learning_rate": 1.612403100775194e-05,
+      "loss": 0.6475,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.018403496545159,
+      "learning_rate": 1.6175710594315245e-05,
+      "loss": 0.7395,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6333717228147262,
+      "learning_rate": 1.6227390180878555e-05,
+      "loss": 0.6498,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6210224917273397,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 0.657,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9938309416887994,
+      "learning_rate": 1.633074935400517e-05,
+      "loss": 0.7661,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8170192545498876,
+      "learning_rate": 1.638242894056848e-05,
+      "loss": 0.6766,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8045935668279347,
+      "learning_rate": 1.6434108527131786e-05,
+      "loss": 0.6594,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8042382015409661,
+      "learning_rate": 1.6485788113695092e-05,
+      "loss": 0.6845,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5535666100814085,
+      "learning_rate": 1.65374677002584e-05,
+      "loss": 0.6375,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6997108185464844,
+      "learning_rate": 1.6589147286821706e-05,
+      "loss": 0.6859,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7942718225337062,
+      "learning_rate": 1.6640826873385016e-05,
+      "loss": 0.7387,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6747243255820246,
+      "learning_rate": 1.6692506459948323e-05,
+      "loss": 0.7287,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6454662283784893,
+      "learning_rate": 1.674418604651163e-05,
+      "loss": 0.6235,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9309376362153705,
+      "learning_rate": 1.6795865633074936e-05,
+      "loss": 0.7316,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7199744977834737,
+      "learning_rate": 1.6847545219638243e-05,
+      "loss": 0.6656,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.860879882125513,
+      "learning_rate": 1.689922480620155e-05,
+      "loss": 0.7451,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6897752046902168,
+      "learning_rate": 1.695090439276486e-05,
+      "loss": 0.6836,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8385185117767382,
+      "learning_rate": 1.7002583979328166e-05,
+      "loss": 0.7228,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7139177537723818,
+      "learning_rate": 1.7054263565891473e-05,
+      "loss": 0.6212,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9065972308984798,
+      "learning_rate": 1.7105943152454783e-05,
+      "loss": 0.736,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.597571037277131,
+      "learning_rate": 1.715762273901809e-05,
+      "loss": 0.7313,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5904093867911444,
+      "learning_rate": 1.7209302325581396e-05,
+      "loss": 0.6848,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7389702620812566,
+      "learning_rate": 1.7260981912144706e-05,
+      "loss": 0.7166,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7843425073020225,
+      "learning_rate": 1.7312661498708013e-05,
+      "loss": 0.7396,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6234645925689897,
+      "learning_rate": 1.736434108527132e-05,
+      "loss": 0.6724,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7248537747976185,
+      "learning_rate": 1.7416020671834626e-05,
+      "loss": 0.6683,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7138102899654486,
+      "learning_rate": 1.7467700258397936e-05,
+      "loss": 0.6655,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9091086293921518,
+      "learning_rate": 1.7519379844961243e-05,
+      "loss": 0.6957,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6963941929137416,
+      "learning_rate": 1.757105943152455e-05,
+      "loss": 0.6796,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7328281549790534,
+      "learning_rate": 1.7622739018087857e-05,
+      "loss": 0.7055,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.589815631589682,
+      "learning_rate": 1.7674418604651163e-05,
+      "loss": 0.7067,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.736638216321355,
+      "learning_rate": 1.772609819121447e-05,
+      "loss": 0.7284,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6392960751987011,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.6809,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5876232640027388,
+      "learning_rate": 1.7829457364341087e-05,
+      "loss": 0.7757,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5829662432851723,
+      "learning_rate": 1.7881136950904393e-05,
+      "loss": 0.7285,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5552442311025254,
+      "learning_rate": 1.79328165374677e-05,
+      "loss": 0.6617,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.888963836541859,
+      "learning_rate": 1.798449612403101e-05,
+      "loss": 0.6744,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7183686266782514,
+      "learning_rate": 1.8036175710594317e-05,
+      "loss": 0.7592,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.754612294162425,
+      "learning_rate": 1.8087855297157624e-05,
+      "loss": 0.7025,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.735550490673282,
+      "learning_rate": 1.813953488372093e-05,
+      "loss": 0.7654,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8993842507501426,
+      "learning_rate": 1.819121447028424e-05,
+      "loss": 0.7315,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7647160578883143,
+      "learning_rate": 1.8242894056847547e-05,
+      "loss": 0.7498,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5622315748437299,
+      "learning_rate": 1.8294573643410854e-05,
+      "loss": 0.6246,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6830378984478318,
+      "learning_rate": 1.8346253229974164e-05,
+      "loss": 0.6808,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6788587788029756,
+      "learning_rate": 1.839793281653747e-05,
+      "loss": 0.6637,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6853859693123143,
+      "learning_rate": 1.8449612403100777e-05,
+      "loss": 0.7021,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7256901120754673,
+      "learning_rate": 1.8501291989664084e-05,
+      "loss": 0.6518,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.693381902243467,
+      "learning_rate": 1.855297157622739e-05,
+      "loss": 0.7455,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.743409280064828,
+      "learning_rate": 1.86046511627907e-05,
+      "loss": 0.7063,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8010066423013003,
+      "learning_rate": 1.8656330749354007e-05,
+      "loss": 0.7634,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7266054450170285,
+      "learning_rate": 1.8708010335917314e-05,
+      "loss": 0.6595,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7604733821001817,
+      "learning_rate": 1.875968992248062e-05,
+      "loss": 0.6977,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5710357215214774,
+      "learning_rate": 1.8811369509043927e-05,
+      "loss": 0.7089,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7716215810245508,
+      "learning_rate": 1.8863049095607237e-05,
+      "loss": 0.6371,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5191226988578574,
+      "learning_rate": 1.8914728682170544e-05,
+      "loss": 0.6987,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6351020494163706,
+      "learning_rate": 1.896640826873385e-05,
+      "loss": 0.7209,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.427810270324823,
+      "learning_rate": 1.9018087855297158e-05,
+      "loss": 0.672,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5550164876229366,
+      "learning_rate": 1.9069767441860468e-05,
+      "loss": 0.7001,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6459538701338634,
+      "learning_rate": 1.9121447028423774e-05,
+      "loss": 0.8093,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6916808859066244,
+      "learning_rate": 1.917312661498708e-05,
+      "loss": 0.8031,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.792743477559064,
+      "learning_rate": 1.922480620155039e-05,
+      "loss": 0.7294,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6043612704787404,
+      "learning_rate": 1.9276485788113698e-05,
+      "loss": 0.6933,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7592731426018393,
+      "learning_rate": 1.9328165374677004e-05,
+      "loss": 0.7709,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6771718193388752,
+      "learning_rate": 1.937984496124031e-05,
+      "loss": 0.6986,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6994093514015336,
+      "learning_rate": 1.943152454780362e-05,
+      "loss": 0.7417,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7915316538248756,
+      "learning_rate": 1.9483204134366928e-05,
+      "loss": 0.7313,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0338506629382898,
+      "learning_rate": 1.9534883720930235e-05,
+      "loss": 0.7696,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.960364695017877,
+      "learning_rate": 1.958656330749354e-05,
+      "loss": 0.7206,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8772124271125357,
+      "learning_rate": 1.9638242894056848e-05,
+      "loss": 0.7129,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9636413049756116,
+      "learning_rate": 1.9689922480620155e-05,
+      "loss": 0.7082,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6527663247417765,
+      "learning_rate": 1.9741602067183465e-05,
+      "loss": 0.6949,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6633158696747927,
+      "learning_rate": 1.979328165374677e-05,
+      "loss": 0.6722,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.721959810193357,
+      "learning_rate": 1.9844961240310078e-05,
+      "loss": 0.7567,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7364744247060135,
+      "learning_rate": 1.9896640826873385e-05,
+      "loss": 0.7176,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7954586915735413,
+      "learning_rate": 1.9948320413436695e-05,
+      "loss": 0.7535,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7346229719620336,
+      "learning_rate": 2e-05,
+      "loss": 0.7229,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7528736498703117,
+      "learning_rate": 1.9999999684324205e-05,
+      "loss": 0.7176,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5477324912304453,
+      "learning_rate": 1.9999998737296837e-05,
+      "loss": 0.6282,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7132943002853476,
+      "learning_rate": 1.9999997158917953e-05,
+      "loss": 0.7502,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6280519729022618,
+      "learning_rate": 1.9999994949187657e-05,
+      "loss": 0.6189,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5954083061779611,
+      "learning_rate": 1.9999992108106083e-05,
+      "loss": 0.7113,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.573154667243791,
+      "learning_rate": 1.9999988635673414e-05,
+      "loss": 0.6588,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7079942025481558,
+      "learning_rate": 1.9999984531889875e-05,
+      "loss": 0.6982,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6879653995321409,
+      "learning_rate": 1.9999979796755715e-05,
+      "loss": 0.666,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.688187141390456,
+      "learning_rate": 1.999997443027124e-05,
+      "loss": 0.6362,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.610216531823565,
+      "learning_rate": 1.9999968432436785e-05,
+      "loss": 0.6575,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3236145772969543,
+      "learning_rate": 1.9999961803252726e-05,
+      "loss": 0.7453,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6347959388024034,
+      "learning_rate": 1.999995454271949e-05,
+      "loss": 0.7709,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6901873630291864,
+      "learning_rate": 1.999994665083753e-05,
+      "loss": 0.7524,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.59903856836198,
+      "learning_rate": 1.9999938127607342e-05,
+      "loss": 0.6619,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5768222530583047,
+      "learning_rate": 1.9999928973029472e-05,
+      "loss": 0.6928,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6138677754701747,
+      "learning_rate": 1.999991918710449e-05,
+      "loss": 0.7036,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.557050816039296,
+      "learning_rate": 1.999990876983302e-05,
+      "loss": 0.6806,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8214136157770862,
+      "learning_rate": 1.999989772121571e-05,
+      "loss": 0.7474,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.776652912399176,
+      "learning_rate": 1.999988604125327e-05,
+      "loss": 0.7904,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6106579305815039,
+      "learning_rate": 1.9999873729946432e-05,
+      "loss": 0.7247,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.614877571449385,
+      "learning_rate": 1.999986078729597e-05,
+      "loss": 0.6763,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5951139560039433,
+      "learning_rate": 1.9999847213302703e-05,
+      "loss": 0.6638,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9395387757454956,
+      "learning_rate": 1.999983300796749e-05,
+      "loss": 0.6751,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.664135711921985,
+      "learning_rate": 1.999981817129123e-05,
+      "loss": 0.7085,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6385503821637333,
+      "learning_rate": 1.9999802703274854e-05,
+      "loss": 0.7145,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.801231013667197,
+      "learning_rate": 1.9999786603919343e-05,
+      "loss": 0.7448,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7969883758602014,
+      "learning_rate": 1.9999769873225706e-05,
+      "loss": 0.7637,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.736104635438498,
+      "learning_rate": 1.999975251119501e-05,
+      "loss": 0.7366,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5335612079356686,
+      "learning_rate": 1.9999734517828345e-05,
+      "loss": 0.6593,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6600035114710792,
+      "learning_rate": 1.999971589312685e-05,
+      "loss": 0.7751,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7123815982846584,
+      "learning_rate": 1.9999696637091698e-05,
+      "loss": 0.7392,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.630631665076908,
+      "learning_rate": 1.9999676749724103e-05,
+      "loss": 0.7332,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6337467422978231,
+      "learning_rate": 1.9999656231025323e-05,
+      "loss": 0.6824,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5854489931811897,
+      "learning_rate": 1.9999635080996655e-05,
+      "loss": 0.698,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.674222207610974,
+      "learning_rate": 1.9999613299639433e-05,
+      "loss": 0.7406,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6331969178252201,
+      "learning_rate": 1.9999590886955033e-05,
+      "loss": 0.7627,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7665659329947683,
+      "learning_rate": 1.999956784294487e-05,
+      "loss": 0.7273,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.642254846355217,
+      "learning_rate": 1.9999544167610396e-05,
+      "loss": 0.6495,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6550105513498534,
+      "learning_rate": 1.999951986095311e-05,
+      "loss": 0.7321,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6336809249086557,
+      "learning_rate": 1.9999494922974544e-05,
+      "loss": 0.7034,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5903316564562422,
+      "learning_rate": 1.9999469353676272e-05,
+      "loss": 0.7146,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.652225354870139,
+      "learning_rate": 1.999944315305991e-05,
+      "loss": 0.7124,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6051610944242776,
+      "learning_rate": 1.999941632112711e-05,
+      "loss": 0.7083,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6624895281927967,
+      "learning_rate": 1.9999388857879568e-05,
+      "loss": 0.6793,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5903397519817577,
+      "learning_rate": 1.9999360763319015e-05,
+      "loss": 0.7418,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6496411308973509,
+      "learning_rate": 1.999933203744723e-05,
+      "loss": 0.699,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.450748921056225,
+      "learning_rate": 1.9999302680266023e-05,
+      "loss": 0.6681,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4417140394303158,
+      "learning_rate": 1.9999272691777246e-05,
+      "loss": 0.6271,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6257541080465774,
+      "learning_rate": 1.99992420719828e-05,
+      "loss": 0.6785,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4910585616627743,
+      "learning_rate": 1.999921082088461e-05,
+      "loss": 0.7389,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6186063135659114,
+      "learning_rate": 1.999917893848465e-05,
+      "loss": 0.6467,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6937318104122674,
+      "learning_rate": 1.9999146424784938e-05,
+      "loss": 0.7253,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5915691938396197,
+      "learning_rate": 1.9999113279787517e-05,
+      "loss": 0.7449,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7866388420647545,
+      "learning_rate": 1.9999079503494496e-05,
+      "loss": 0.7856,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6184684361257564,
+      "learning_rate": 1.9999045095907988e-05,
+      "loss": 0.7174,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3692111923599486,
+      "learning_rate": 1.9999010057030183e-05,
+      "loss": 0.661,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.647713814205106,
+      "learning_rate": 1.9998974386863276e-05,
+      "loss": 0.6867,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9018199285631465,
+      "learning_rate": 1.9998938085409534e-05,
+      "loss": 0.731,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.562980883031733,
+      "learning_rate": 1.9998901152671243e-05,
+      "loss": 0.6987,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6468384566767702,
+      "learning_rate": 1.9998863588650732e-05,
+      "loss": 0.7059,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7768572801799298,
+      "learning_rate": 1.9998825393350375e-05,
+      "loss": 0.7415,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6625920635919154,
+      "learning_rate": 1.999878656677259e-05,
+      "loss": 0.7469,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6094740309384488,
+      "learning_rate": 1.9998747108919815e-05,
+      "loss": 0.6855,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.991797913997878,
+      "learning_rate": 1.999870701979455e-05,
+      "loss": 0.7273,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7667346269503605,
+      "learning_rate": 1.9998666299399326e-05,
+      "loss": 0.7284,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.766153239672727,
+      "learning_rate": 1.999862494773671e-05,
+      "loss": 0.7407,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.656801599742351,
+      "learning_rate": 1.9998582964809317e-05,
+      "loss": 0.7124,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7539647693684355,
+      "learning_rate": 1.9998540350619793e-05,
+      "loss": 0.7933,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.668483824343145,
+      "learning_rate": 1.9998497105170833e-05,
+      "loss": 0.6857,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7266829090508202,
+      "learning_rate": 1.9998453228465165e-05,
+      "loss": 0.7264,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.504589213075702,
+      "learning_rate": 1.999840872050556e-05,
+      "loss": 0.6976,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5411625201048247,
+      "learning_rate": 1.999836358129483e-05,
+      "loss": 0.6803,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6495119182136446,
+      "learning_rate": 1.9998317810835815e-05,
+      "loss": 0.6846,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5180157554522513,
+      "learning_rate": 1.9998271409131417e-05,
+      "loss": 0.7037,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7127400835580084,
+      "learning_rate": 1.999822437618456e-05,
+      "loss": 0.7223,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.608780167659147,
+      "learning_rate": 1.9998176711998215e-05,
+      "loss": 0.7638,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5538310083092766,
+      "learning_rate": 1.999812841657539e-05,
+      "loss": 0.7472,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6357233849199189,
+      "learning_rate": 1.9998079489919134e-05,
+      "loss": 0.7257,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5671861186454064,
+      "learning_rate": 1.9998029932032534e-05,
+      "loss": 0.6688,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5890134850728637,
+      "learning_rate": 1.9997979742918723e-05,
+      "loss": 0.7639,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5196065614370386,
+      "learning_rate": 1.999792892258087e-05,
+      "loss": 0.6138,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5383873247623239,
+      "learning_rate": 1.9997877471022182e-05,
+      "loss": 0.6591,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.645732015396448,
+      "learning_rate": 1.9997825388245905e-05,
+      "loss": 0.7215,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.524508212479353,
+      "learning_rate": 1.999777267425533e-05,
+      "loss": 0.6901,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.481353656344975,
+      "learning_rate": 1.9997719329053782e-05,
+      "loss": 0.7021,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6202897985297207,
+      "learning_rate": 1.999766535264463e-05,
+      "loss": 0.7857,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5237698705940714,
+      "learning_rate": 1.9997610745031292e-05,
+      "loss": 0.6699,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6958931486685858,
+      "learning_rate": 1.99975555062172e-05,
+      "loss": 0.7847,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8245844668413667,
+      "learning_rate": 1.9997499636205847e-05,
+      "loss": 0.7416,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7432899579231305,
+      "learning_rate": 1.9997443135000765e-05,
+      "loss": 0.7082,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6527655313431695,
+      "learning_rate": 1.9997386002605515e-05,
+      "loss": 0.7221,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.526372576391039,
+      "learning_rate": 1.999732823902371e-05,
+      "loss": 0.7254,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.50463548295059,
+      "learning_rate": 1.9997269844258993e-05,
+      "loss": 0.6663,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5833940828949773,
+      "learning_rate": 1.999721081831505e-05,
+      "loss": 0.6769,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5093683309782873,
+      "learning_rate": 1.9997151161195613e-05,
+      "loss": 0.6292,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5451827377593428,
+      "learning_rate": 1.9997090872904442e-05,
+      "loss": 0.7566,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.652419356832569,
+      "learning_rate": 1.9997029953445345e-05,
+      "loss": 0.6829,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6704165791129075,
+      "learning_rate": 1.999696840282217e-05,
+      "loss": 0.678,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6164923754204596,
+      "learning_rate": 1.9996906221038802e-05,
+      "loss": 0.7532,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5693936672642468,
+      "learning_rate": 1.999684340809917e-05,
+      "loss": 0.6923,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.678340069789809,
+      "learning_rate": 1.9996779964007232e-05,
+      "loss": 0.6951,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4737192056559285,
+      "learning_rate": 1.9996715888767e-05,
+      "loss": 0.6884,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4676956288112966,
+      "learning_rate": 1.9996651182382518e-05,
+      "loss": 0.7117,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4944319696547814,
+      "learning_rate": 1.999658584485787e-05,
+      "loss": 0.6733,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5298826379579882,
+      "learning_rate": 1.9996519876197185e-05,
+      "loss": 0.7322,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5099953776408808,
+      "learning_rate": 1.999645327640462e-05,
+      "loss": 0.7466,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4888009192841576,
+      "learning_rate": 1.999638604548439e-05,
+      "loss": 0.725,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4929981526224776,
+      "learning_rate": 1.9996318183440732e-05,
+      "loss": 0.7063,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.598305493345369,
+      "learning_rate": 1.9996249690277934e-05,
+      "loss": 0.7423,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4226248304572933,
+      "learning_rate": 1.999618056600032e-05,
+      "loss": 0.6704,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7489322402155776,
+      "learning_rate": 1.999611081061225e-05,
+      "loss": 0.725,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5527406704373163,
+      "learning_rate": 1.999604042411813e-05,
+      "loss": 0.6761,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.769345305317264,
+      "learning_rate": 1.9995969406522412e-05,
+      "loss": 0.7771,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6652858338714835,
+      "learning_rate": 1.9995897757829564e-05,
+      "loss": 0.7055,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.663801217498132,
+      "learning_rate": 1.9995825478044126e-05,
+      "loss": 0.7385,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7078167048871067,
+      "learning_rate": 1.999575256717065e-05,
+      "loss": 0.7001,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5668344269206198,
+      "learning_rate": 1.9995679025213747e-05,
+      "loss": 0.6665,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6364168239147343,
+      "learning_rate": 1.9995604852178055e-05,
+      "loss": 0.7096,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5765966436282393,
+      "learning_rate": 1.9995530048068253e-05,
+      "loss": 0.7309,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5192936807178008,
+      "learning_rate": 1.9995454612889076e-05,
+      "loss": 0.6978,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6293084908077977,
+      "learning_rate": 1.9995378546645274e-05,
+      "loss": 0.6854,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.667735456460159,
+      "learning_rate": 1.999530184934166e-05,
+      "loss": 0.7553,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5729917377805436,
+      "learning_rate": 1.9995224520983068e-05,
+      "loss": 0.6877,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.564234495661959,
+      "learning_rate": 1.9995146561574384e-05,
+      "loss": 0.6676,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5438824616101234,
+      "learning_rate": 1.9995067971120527e-05,
+      "loss": 0.6212,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8259770743360872,
+      "learning_rate": 1.9994988749626464e-05,
+      "loss": 0.702,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5362335385464132,
+      "learning_rate": 1.9994908897097195e-05,
+      "loss": 0.7317,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6953742372021292,
+      "learning_rate": 1.999482841353776e-05,
+      "loss": 0.7131,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6076065272604732,
+      "learning_rate": 1.999474729895324e-05,
+      "loss": 0.7146,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.477019869898655,
+      "learning_rate": 1.9994665553348757e-05,
+      "loss": 0.6224,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5236802910247287,
+      "learning_rate": 1.999458317672947e-05,
+      "loss": 0.633,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4541173028446257,
+      "learning_rate": 1.9994500169100583e-05,
+      "loss": 0.7029,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4766634951974353,
+      "learning_rate": 1.9994416530467336e-05,
+      "loss": 0.6331,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.688122387304618,
+      "learning_rate": 1.9994332260835007e-05,
+      "loss": 0.7227,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4646164374410906,
+      "learning_rate": 1.9994247360208924e-05,
+      "loss": 0.675,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5011274551065308,
+      "learning_rate": 1.9994161828594435e-05,
+      "loss": 0.6439,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6908090431319895,
+      "learning_rate": 1.9994075665996952e-05,
+      "loss": 0.6839,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5981954023945975,
+      "learning_rate": 1.9993988872421902e-05,
+      "loss": 0.7136,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5478585709374062,
+      "learning_rate": 1.999390144787478e-05,
+      "loss": 0.7035,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4579428104492471,
+      "learning_rate": 1.9993813392361095e-05,
+      "loss": 0.6551,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5187776947636051,
+      "learning_rate": 1.999372470588641e-05,
+      "loss": 0.6541,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4904409842799748,
+      "learning_rate": 1.999363538845632e-05,
+      "loss": 0.6434,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.619864565403804,
+      "learning_rate": 1.9993545440076473e-05,
+      "loss": 0.7437,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6021414105917995,
+      "learning_rate": 1.9993454860752538e-05,
+      "loss": 0.6774,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5584038257804533,
+      "learning_rate": 1.999336365049024e-05,
+      "loss": 0.6537,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5625273892863119,
+      "learning_rate": 1.9993271809295337e-05,
+      "loss": 0.7039,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4000568497559387,
+      "learning_rate": 1.9993179337173624e-05,
+      "loss": 0.6784,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.828352824303991,
+      "learning_rate": 1.9993086234130944e-05,
+      "loss": 0.732,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6990779423744238,
+      "learning_rate": 1.999299250017317e-05,
+      "loss": 0.6641,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.604924316166706,
+      "learning_rate": 1.9992898135306223e-05,
+      "loss": 0.7196,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6965750638194799,
+      "learning_rate": 1.999280313953606e-05,
+      "loss": 0.6855,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5202492241075667,
+      "learning_rate": 1.999270751286868e-05,
+      "loss": 0.6212,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6066269606647534,
+      "learning_rate": 1.9992611255310115e-05,
+      "loss": 0.7084,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6071333597296182,
+      "learning_rate": 1.9992514366866453e-05,
+      "loss": 0.6485,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5036425867189884,
+      "learning_rate": 1.9992416847543802e-05,
+      "loss": 0.7262,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5939623373361334,
+      "learning_rate": 1.9992318697348318e-05,
+      "loss": 0.6811,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8675489514738624,
+      "learning_rate": 1.9992219916286205e-05,
+      "loss": 0.7211,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.497878083787526,
+      "learning_rate": 1.9992120504363694e-05,
+      "loss": 0.6811,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4579373321565814,
+      "learning_rate": 1.9992020461587063e-05,
+      "loss": 0.6915,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.591533990226762,
+      "learning_rate": 1.9991919787962627e-05,
+      "loss": 0.6991,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5554914839734313,
+      "learning_rate": 1.9991818483496747e-05,
+      "loss": 0.6941,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.587951508841042,
+      "learning_rate": 1.999171654819581e-05,
+      "loss": 0.667,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4680807333841988,
+      "learning_rate": 1.999161398206626e-05,
+      "loss": 0.6474,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5560024968546469,
+      "learning_rate": 1.999151078511457e-05,
+      "loss": 0.6836,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6578572949425674,
+      "learning_rate": 1.999140695734725e-05,
+      "loss": 0.6271,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5658728245412266,
+      "learning_rate": 1.9991302498770867e-05,
+      "loss": 0.7176,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5784429097767325,
+      "learning_rate": 1.9991197409392004e-05,
+      "loss": 0.6789,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6372971707589818,
+      "learning_rate": 1.9991091689217303e-05,
+      "loss": 0.703,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.421081573115393,
+      "learning_rate": 1.9990985338253434e-05,
+      "loss": 0.7214,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5860319062495565,
+      "learning_rate": 1.9990878356507116e-05,
+      "loss": 0.6939,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6161604121398174,
+      "learning_rate": 1.99907707439851e-05,
+      "loss": 0.7171,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.587284662254865,
+      "learning_rate": 1.9990662500694183e-05,
+      "loss": 0.7268,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5060141319381881,
+      "learning_rate": 1.9990553626641194e-05,
+      "loss": 0.698,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4562455173145112,
+      "learning_rate": 1.9990444121833016e-05,
+      "loss": 0.658,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4167032611084793,
+      "learning_rate": 1.9990333986276552e-05,
+      "loss": 0.6858,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5886033676755311,
+      "learning_rate": 1.999022321997876e-05,
+      "loss": 0.6974,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5288863331350755,
+      "learning_rate": 1.9990111822946634e-05,
+      "loss": 0.6808,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2469602818591705,
+      "learning_rate": 1.9989999795187206e-05,
+      "loss": 0.6634,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4832806217100252,
+      "learning_rate": 1.998988713670755e-05,
+      "loss": 0.6709,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5277468254344768,
+      "learning_rate": 1.998977384751478e-05,
+      "loss": 0.7142,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5811186996046123,
+      "learning_rate": 1.9989659927616044e-05,
+      "loss": 0.7135,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6281699293384213,
+      "learning_rate": 1.9989545377018538e-05,
+      "loss": 0.6752,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5158983750917991,
+      "learning_rate": 1.9989430195729494e-05,
+      "loss": 0.6805,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6764062814021865,
+      "learning_rate": 1.998931438375618e-05,
+      "loss": 0.7161,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4806756753167973,
+      "learning_rate": 1.998919794110591e-05,
+      "loss": 0.7263,
+      "step": 572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4557835859341586,
+      "learning_rate": 1.998908086778604e-05,
+      "loss": 0.7002,
+      "step": 573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5920415959315477,
+      "learning_rate": 1.9988963163803958e-05,
+      "loss": 0.7437,
+      "step": 574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5396736596865324,
+      "learning_rate": 1.9988844829167092e-05,
+      "loss": 0.6971,
+      "step": 575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6295908854475456,
+      "learning_rate": 1.9988725863882922e-05,
+      "loss": 0.7321,
+      "step": 576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4903184459565266,
+      "learning_rate": 1.998860626795895e-05,
+      "loss": 0.69,
+      "step": 577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5427135401330314,
+      "learning_rate": 1.998848604140273e-05,
+      "loss": 0.6935,
+      "step": 578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4066609100175793,
+      "learning_rate": 1.998836518422185e-05,
+      "loss": 0.6507,
+      "step": 579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5534434492894746,
+      "learning_rate": 1.9988243696423947e-05,
+      "loss": 0.6572,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5255378525100314,
+      "learning_rate": 1.9988121578016683e-05,
+      "loss": 0.693,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7531413086049314,
+      "learning_rate": 1.9987998829007775e-05,
+      "loss": 0.7463,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.501387987108143,
+      "learning_rate": 1.9987875449404965e-05,
+      "loss": 0.6806,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5181521553761232,
+      "learning_rate": 1.998775143921605e-05,
+      "loss": 0.69,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7491486385822463,
+      "learning_rate": 1.9987626798448858e-05,
+      "loss": 0.6683,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5476621686426533,
+      "learning_rate": 1.9987501527111253e-05,
+      "loss": 0.6665,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5106475741270504,
+      "learning_rate": 1.9987375625211155e-05,
+      "loss": 0.6797,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5416137926811995,
+      "learning_rate": 1.99872490927565e-05,
+      "loss": 0.6635,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5484727256917714,
+      "learning_rate": 1.9987121929755284e-05,
+      "loss": 0.6813,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4053969444944152,
+      "learning_rate": 1.9986994136215533e-05,
+      "loss": 0.6086,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4333779487023979,
+      "learning_rate": 1.9986865712145316e-05,
+      "loss": 0.6764,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4560090025214762,
+      "learning_rate": 1.9986736657552742e-05,
+      "loss": 0.6517,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.267194743443912,
+      "learning_rate": 1.9986606972445956e-05,
+      "loss": 0.7425,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.579606748605446,
+      "learning_rate": 1.998647665683315e-05,
+      "loss": 0.7427,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5422188397192407,
+      "learning_rate": 1.998634571072255e-05,
+      "loss": 0.701,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6001968143530119,
+      "learning_rate": 1.998621413412242e-05,
+      "loss": 0.6989,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5119614517202604,
+      "learning_rate": 1.998608192704107e-05,
+      "loss": 0.6976,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.607026098945014,
+      "learning_rate": 1.9985949089486847e-05,
+      "loss": 0.6309,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4701575981050437,
+      "learning_rate": 1.998581562146814e-05,
+      "loss": 0.7231,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6280420878966146,
+      "learning_rate": 1.998568152299337e-05,
+      "loss": 0.7428,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3699178843009088,
+      "learning_rate": 1.9985546794071006e-05,
+      "loss": 0.6128,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5123627151102557,
+      "learning_rate": 1.9985411434709553e-05,
+      "loss": 0.6496,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4369175601743995,
+      "learning_rate": 1.998527544491756e-05,
+      "loss": 0.6871,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5328196051974403,
+      "learning_rate": 1.998513882470361e-05,
+      "loss": 0.6637,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4953447264703434,
+      "learning_rate": 1.998500157407633e-05,
+      "loss": 0.6365,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4521188841982653,
+      "learning_rate": 1.9984863693044385e-05,
+      "loss": 0.6142,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5928050212897833,
+      "learning_rate": 1.998472518161648e-05,
+      "loss": 0.7354,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5919485944279186,
+      "learning_rate": 1.998458603980136e-05,
+      "loss": 0.7091,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3083587891031876,
+      "learning_rate": 1.998444626760781e-05,
+      "loss": 0.6477,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3466313984866822,
+      "learning_rate": 1.9984305865044654e-05,
+      "loss": 0.6531,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5911592107315577,
+      "learning_rate": 1.9984164832120755e-05,
+      "loss": 0.7384,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6353324900870023,
+      "learning_rate": 1.998402316884502e-05,
+      "loss": 0.7327,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.653308631779101,
+      "learning_rate": 1.998388087522639e-05,
+      "loss": 0.671,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4320412526715238,
+      "learning_rate": 1.9983737951273854e-05,
+      "loss": 0.6804,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5341690422480376,
+      "learning_rate": 1.9983594396996428e-05,
+      "loss": 0.705,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6323730023277883,
+      "learning_rate": 1.998345021240318e-05,
+      "loss": 0.6864,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5669000089982936,
+      "learning_rate": 1.9983305397503214e-05,
+      "loss": 0.636,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.501495490032769,
+      "learning_rate": 1.9983159952305668e-05,
+      "loss": 0.6606,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5710918712936695,
+      "learning_rate": 1.998301387681973e-05,
+      "loss": 0.64,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4704143752378123,
+      "learning_rate": 1.9982867171054622e-05,
+      "loss": 0.6388,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5505054142084593,
+      "learning_rate": 1.9982719835019604e-05,
+      "loss": 0.7019,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5782432889513178,
+      "learning_rate": 1.9982571868723975e-05,
+      "loss": 0.7124,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5165856846299397,
+      "learning_rate": 1.9982423272177087e-05,
+      "loss": 0.6638,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.621152872133489,
+      "learning_rate": 1.998227404538831e-05,
+      "loss": 0.7407,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5111365170462379,
+      "learning_rate": 1.998212418836707e-05,
+      "loss": 0.7225,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5483996651316276,
+      "learning_rate": 1.998197370112283e-05,
+      "loss": 0.7263,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.399157163115521,
+      "learning_rate": 1.9981822583665094e-05,
+      "loss": 0.6447,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4471568943214923,
+      "learning_rate": 1.9981670836003396e-05,
+      "loss": 0.6718,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4901968575050888,
+      "learning_rate": 1.998151845814732e-05,
+      "loss": 0.6279,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5009167571801751,
+      "learning_rate": 1.9981365450106484e-05,
+      "loss": 0.6916,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.597806816187955,
+      "learning_rate": 1.9981211811890554e-05,
+      "loss": 0.6974,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.576796548519489,
+      "learning_rate": 1.998105754350922e-05,
+      "loss": 0.7309,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5228990434634961,
+      "learning_rate": 1.9980902644972234e-05,
+      "loss": 0.7332,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5724283264516041,
+      "learning_rate": 1.998074711628937e-05,
+      "loss": 0.6653,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5764066197560027,
+      "learning_rate": 1.9980590957470437e-05,
+      "loss": 0.6611,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.624152035648111,
+      "learning_rate": 1.9980434168525315e-05,
+      "loss": 0.6929,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4293267467178583,
+      "learning_rate": 1.9980276749463886e-05,
+      "loss": 0.6679,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5297459592159572,
+      "learning_rate": 1.9980118700296095e-05,
+      "loss": 0.647,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5407633411746564,
+      "learning_rate": 1.997996002103192e-05,
+      "loss": 0.6684,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4553574671565237,
+      "learning_rate": 1.997980071168138e-05,
+      "loss": 0.6923,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5706140788465575,
+      "learning_rate": 1.9979640772254528e-05,
+      "loss": 0.7113,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.500585441782913,
+      "learning_rate": 1.997948020276147e-05,
+      "loss": 0.6686,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6194846382554953,
+      "learning_rate": 1.9979319003212337e-05,
+      "loss": 0.696,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6542761214754882,
+      "learning_rate": 1.997915717361731e-05,
+      "loss": 0.7594,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5234030695227856,
+      "learning_rate": 1.9978994713986606e-05,
+      "loss": 0.7039,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5943618048911075,
+      "learning_rate": 1.9978831624330483e-05,
+      "loss": 0.6635,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4703588399364698,
+      "learning_rate": 1.997866790465923e-05,
+      "loss": 0.7,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.484639435353238,
+      "learning_rate": 1.997850355498319e-05,
+      "loss": 0.6823,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4989152800964898,
+      "learning_rate": 1.9978338575312742e-05,
+      "loss": 0.7432,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4212172108057681,
+      "learning_rate": 1.9978172965658297e-05,
+      "loss": 0.6346,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5397759347630786,
+      "learning_rate": 1.997800672603031e-05,
+      "loss": 0.7287,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.471766681289549,
+      "learning_rate": 1.997783985643928e-05,
+      "loss": 0.6427,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.406277168859184,
+      "learning_rate": 1.997767235689574e-05,
+      "loss": 0.6361,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.540609366625842,
+      "learning_rate": 1.9977504227410268e-05,
+      "loss": 0.7265,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4956908318409674,
+      "learning_rate": 1.9977335467993474e-05,
+      "loss": 0.7131,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7909151171788398,
+      "learning_rate": 1.997716607865602e-05,
+      "loss": 0.6775,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7897552082035544,
+      "learning_rate": 1.9976996059408595e-05,
+      "loss": 0.7351,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4937434048686618,
+      "learning_rate": 1.997682541026193e-05,
+      "loss": 0.657,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5901092382609006,
+      "learning_rate": 1.997665413122681e-05,
+      "loss": 0.7749,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4276446980963735,
+      "learning_rate": 1.9976482222314034e-05,
+      "loss": 0.7159,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5174070283919316,
+      "learning_rate": 1.997630968353447e-05,
+      "loss": 0.6609,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.586215741784857,
+      "learning_rate": 1.9976136514899e-05,
+      "loss": 0.6905,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5148969651741977,
+      "learning_rate": 1.9975962716418565e-05,
+      "loss": 0.7307,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4337425050813954,
+      "learning_rate": 1.9975788288104132e-05,
+      "loss": 0.712,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3863645719164037,
+      "learning_rate": 1.997561322996672e-05,
+      "loss": 0.643,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6304058974927758,
+      "learning_rate": 1.9975437542017372e-05,
+      "loss": 0.7391,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5533666319136317,
+      "learning_rate": 1.9975261224267187e-05,
+      "loss": 0.6778,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6759539606330542,
+      "learning_rate": 1.9975084276727298e-05,
+      "loss": 0.812,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5050305412209861,
+      "learning_rate": 1.9974906699408874e-05,
+      "loss": 0.6326,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.428899912874864,
+      "learning_rate": 1.9974728492323122e-05,
+      "loss": 0.6581,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6176892667761549,
+      "learning_rate": 1.9974549655481302e-05,
+      "loss": 0.7051,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6651281249090957,
+      "learning_rate": 1.9974370188894698e-05,
+      "loss": 0.745,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4779479777687863,
+      "learning_rate": 1.9974190092574645e-05,
+      "loss": 0.6708,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4294174856552997,
+      "learning_rate": 1.997400936653251e-05,
+      "loss": 0.6468,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.585851807933539,
+      "learning_rate": 1.9973828010779702e-05,
+      "loss": 0.7352,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5257996073144429,
+      "learning_rate": 1.9973646025327678e-05,
+      "loss": 0.686,
+      "step": 676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4566771048747251,
+      "learning_rate": 1.9973463410187922e-05,
+      "loss": 0.6547,
+      "step": 677
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4923679582215776,
+      "learning_rate": 1.9973280165371964e-05,
+      "loss": 0.7021,
+      "step": 678
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.619787806989808,
+      "learning_rate": 1.9973096290891374e-05,
+      "loss": 0.756,
+      "step": 679
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5322144254418917,
+      "learning_rate": 1.997291178675776e-05,
+      "loss": 0.6807,
+      "step": 680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4520212719360082,
+      "learning_rate": 1.997272665298277e-05,
+      "loss": 0.6949,
+      "step": 681
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3920593151329614,
+      "learning_rate": 1.99725408895781e-05,
+      "loss": 0.6489,
+      "step": 682
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.383893468736846,
+      "learning_rate": 1.9972354496555467e-05,
+      "loss": 0.7454,
+      "step": 683
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5615148872107998,
+      "learning_rate": 1.997216747392664e-05,
+      "loss": 0.7026,
+      "step": 684
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4900843793543455,
+      "learning_rate": 1.9971979821703437e-05,
+      "loss": 0.7047,
+      "step": 685
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4819541062470392,
+      "learning_rate": 1.99717915398977e-05,
+      "loss": 0.6941,
+      "step": 686
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4353172481897687,
+      "learning_rate": 1.9971602628521312e-05,
+      "loss": 0.6921,
+      "step": 687
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4927750474281518,
+      "learning_rate": 1.9971413087586207e-05,
+      "loss": 0.6746,
+      "step": 688
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5006879182754358,
+      "learning_rate": 1.9971222917104344e-05,
+      "loss": 0.6387,
+      "step": 689
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5965338684131045,
+      "learning_rate": 1.9971032117087736e-05,
+      "loss": 0.6591,
+      "step": 690
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5332447216441298,
+      "learning_rate": 1.9970840687548425e-05,
+      "loss": 0.6956,
+      "step": 691
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5038912685688457,
+      "learning_rate": 1.99706486284985e-05,
+      "loss": 0.7015,
+      "step": 692
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3650321756140882,
+      "learning_rate": 1.9970455939950085e-05,
+      "loss": 0.6381,
+      "step": 693
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.449135667419262,
+      "learning_rate": 1.9970262621915348e-05,
+      "loss": 0.6802,
+      "step": 694
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.511772371666896,
+      "learning_rate": 1.9970068674406487e-05,
+      "loss": 0.7023,
+      "step": 695
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.403623331865085,
+      "learning_rate": 1.9969874097435754e-05,
+      "loss": 0.6135,
+      "step": 696
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6073687001143515,
+      "learning_rate": 1.996967889101543e-05,
+      "loss": 0.6853,
+      "step": 697
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4207683910097149,
+      "learning_rate": 1.9969483055157846e-05,
+      "loss": 0.6793,
+      "step": 698
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5201503404503816,
+      "learning_rate": 1.9969286589875358e-05,
+      "loss": 0.6931,
+      "step": 699
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.413567050526946,
+      "learning_rate": 1.9969089495180372e-05,
+      "loss": 0.7178,
+      "step": 700
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.458848571585027,
+      "learning_rate": 1.9968891771085334e-05,
+      "loss": 0.6818,
+      "step": 701
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5337122369546061,
+      "learning_rate": 1.996869341760272e-05,
+      "loss": 0.765,
+      "step": 702
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.423220909214009,
+      "learning_rate": 1.9968494434745065e-05,
+      "loss": 0.6991,
+      "step": 703
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4333970769118678,
+      "learning_rate": 1.9968294822524923e-05,
+      "loss": 0.6886,
+      "step": 704
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4869486933144944,
+      "learning_rate": 1.99680945809549e-05,
+      "loss": 0.7294,
+      "step": 705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.447119989931308,
+      "learning_rate": 1.9967893710047638e-05,
+      "loss": 0.6627,
+      "step": 706
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.54846110090175,
+      "learning_rate": 1.9967692209815818e-05,
+      "loss": 0.7537,
+      "step": 707
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.429511220935334,
+      "learning_rate": 1.996749008027216e-05,
+      "loss": 0.6193,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5018334310035786,
+      "learning_rate": 1.9967287321429432e-05,
+      "loss": 0.6831,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5922078918409035,
+      "learning_rate": 1.996708393330043e-05,
+      "loss": 0.7098,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4807718012098625,
+      "learning_rate": 1.9966879915897996e-05,
+      "loss": 0.6938,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4376444536688462,
+      "learning_rate": 1.996667526923501e-05,
+      "loss": 0.7138,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3909481830310941,
+      "learning_rate": 1.996646999332439e-05,
+      "loss": 0.662,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4544940365026182,
+      "learning_rate": 1.9966264088179105e-05,
+      "loss": 0.6851,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3572577288062038,
+      "learning_rate": 1.9966057553812144e-05,
+      "loss": 0.7007,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.496013748778949,
+      "learning_rate": 1.9965850390236554e-05,
+      "loss": 0.6698,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.488084995416177,
+      "learning_rate": 1.9965642597465412e-05,
+      "loss": 0.7788,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4330696166458783,
+      "learning_rate": 1.9965434175511837e-05,
+      "loss": 0.6799,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4924156453944606,
+      "learning_rate": 1.9965225124388982e-05,
+      "loss": 0.6159,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.533901488044266,
+      "learning_rate": 1.9965015444110058e-05,
+      "loss": 0.6911,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5045640175797126,
+      "learning_rate": 1.9964805134688294e-05,
+      "loss": 0.6553,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.365117451432009,
+      "learning_rate": 1.996459419613697e-05,
+      "loss": 0.6434,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.48307486364579,
+      "learning_rate": 1.9964382628469403e-05,
+      "loss": 0.6738,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4180241513529275,
+      "learning_rate": 1.9964170431698953e-05,
+      "loss": 0.6418,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.545127420942442,
+      "learning_rate": 1.9963957605839014e-05,
+      "loss": 0.7338,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4687238650330188,
+      "learning_rate": 1.9963744150903026e-05,
+      "loss": 0.6834,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.47330215819562,
+      "learning_rate": 1.996353006690446e-05,
+      "loss": 0.7379,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.319754126942464,
+      "learning_rate": 1.996331535385684e-05,
+      "loss": 0.6928,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3881803734601454,
+      "learning_rate": 1.9963100011773716e-05,
+      "loss": 0.6286,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5061512230106704,
+      "learning_rate": 1.9962884040668686e-05,
+      "loss": 0.6475,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4580290704034682,
+      "learning_rate": 1.9962667440555383e-05,
+      "loss": 0.6568,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5634361514433177,
+      "learning_rate": 1.9962450211447485e-05,
+      "loss": 0.7047,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.478199450047416,
+      "learning_rate": 1.9962232353358707e-05,
+      "loss": 0.6478,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.433620691695921,
+      "learning_rate": 1.99620138663028e-05,
+      "loss": 0.6053,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5141799172616108,
+      "learning_rate": 1.9961794750293558e-05,
+      "loss": 0.6871,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5598473729761482,
+      "learning_rate": 1.9961575005344822e-05,
+      "loss": 0.7024,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3240783076797853,
+      "learning_rate": 1.996135463147046e-05,
+      "loss": 0.6251,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.350872767863367,
+      "learning_rate": 1.9961133628684382e-05,
+      "loss": 0.6219,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6001908546105366,
+      "learning_rate": 1.996091199700055e-05,
+      "loss": 0.7471,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3954985012974443,
+      "learning_rate": 1.9960689736432952e-05,
+      "loss": 0.6987,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.412274109661438,
+      "learning_rate": 1.996046684699562e-05,
+      "loss": 0.672,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5066987504890756,
+      "learning_rate": 1.9960243328702628e-05,
+      "loss": 0.6547,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4124857083163054,
+      "learning_rate": 1.9960019181568082e-05,
+      "loss": 0.6698,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.397114388863728,
+      "learning_rate": 1.995979440560614e-05,
+      "loss": 0.7203,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.417821030501763,
+      "learning_rate": 1.9959569000830993e-05,
+      "loss": 0.7162,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.409900420169347,
+      "learning_rate": 1.995934296725687e-05,
+      "loss": 0.6632,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5140070980868126,
+      "learning_rate": 1.9959116304898045e-05,
+      "loss": 0.7177,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5410523694278575,
+      "learning_rate": 1.995888901376882e-05,
+      "loss": 0.6075,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4903310841869246,
+      "learning_rate": 1.9958661093883552e-05,
+      "loss": 0.6364,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5639198718362435,
+      "learning_rate": 1.995843254525663e-05,
+      "loss": 0.7242,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4392599858767008,
+      "learning_rate": 1.9958203367902482e-05,
+      "loss": 0.657,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.372316950558404,
+      "learning_rate": 1.995797356183558e-05,
+      "loss": 0.6053,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4731749574770556,
+      "learning_rate": 1.9957743127070427e-05,
+      "loss": 0.6608,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5437065581722909,
+      "learning_rate": 1.995751206362158e-05,
+      "loss": 0.6605,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4941770062972664,
+      "learning_rate": 1.9957280371503617e-05,
+      "loss": 0.7149,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5275310586871569,
+      "learning_rate": 1.9957048050731175e-05,
+      "loss": 0.6705,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3402932972687553,
+      "learning_rate": 1.9956815101318916e-05,
+      "loss": 0.6265,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4207912968358263,
+      "learning_rate": 1.995658152328155e-05,
+      "loss": 0.708,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4574596603595302,
+      "learning_rate": 1.9956347316633824e-05,
+      "loss": 0.6417,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3379912178481204,
+      "learning_rate": 1.995611248139052e-05,
+      "loss": 0.6392,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4350808561453696,
+      "learning_rate": 1.995587701756647e-05,
+      "loss": 0.6429,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4937766834892385,
+      "learning_rate": 1.9955640925176543e-05,
+      "loss": 0.7354,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.457003268466653,
+      "learning_rate": 1.995540420423564e-05,
+      "loss": 0.6066,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5086050371945936,
+      "learning_rate": 1.99551668547587e-05,
+      "loss": 0.7269,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.578153931003182,
+      "learning_rate": 1.9954928876760718e-05,
+      "loss": 0.7003,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5626557844703117,
+      "learning_rate": 1.9954690270256717e-05,
+      "loss": 0.741,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3792139417876308,
+      "learning_rate": 1.9954451035261754e-05,
+      "loss": 0.6957,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4373695480199493,
+      "learning_rate": 1.9954211171790946e-05,
+      "loss": 0.6691,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5044349434688922,
+      "learning_rate": 1.9953970679859425e-05,
+      "loss": 0.7092,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4607424682339447,
+      "learning_rate": 1.9953729559482383e-05,
+      "loss": 0.6683,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4423507449253523,
+      "learning_rate": 1.9953487810675036e-05,
+      "loss": 0.7058,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2487904895866315,
+      "learning_rate": 1.995324543345265e-05,
+      "loss": 0.6222,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.401285308678884,
+      "learning_rate": 1.995300242783053e-05,
+      "loss": 0.6654,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3711359609860783,
+      "learning_rate": 1.9952758793824016e-05,
+      "loss": 0.6917,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5597283767544046,
+      "learning_rate": 1.995251453144849e-05,
+      "loss": 0.7009,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.412779674017897,
+      "learning_rate": 1.995226964071937e-05,
+      "loss": 0.7018,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3907166782976859,
+      "learning_rate": 1.9952024121652122e-05,
+      "loss": 0.6661,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3705489769068018,
+      "learning_rate": 1.9951777974262247e-05,
+      "loss": 0.7182,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4892289956228386,
+      "learning_rate": 1.9951531198565287e-05,
+      "loss": 0.7256,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4639856698997271,
+      "learning_rate": 1.9951283794576814e-05,
+      "loss": 0.6965,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.421904008171216,
+      "learning_rate": 1.9951035762312453e-05,
+      "loss": 0.6684,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4630548973187436,
+      "learning_rate": 1.995078710178787e-05,
+      "loss": 0.7054,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4092327274484135,
+      "learning_rate": 1.9950537813018753e-05,
+      "loss": 0.6435,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5645408272419137,
+      "learning_rate": 1.9950287896020846e-05,
+      "loss": 0.6852,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4678855949590572,
+      "learning_rate": 1.995003735080993e-05,
+      "loss": 0.7105,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5484468584903797,
+      "learning_rate": 1.9949786177401816e-05,
+      "loss": 0.7102,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4400869394919489,
+      "learning_rate": 1.994953437581237e-05,
+      "loss": 0.6719,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5901981493001038,
+      "learning_rate": 1.9949281946057482e-05,
+      "loss": 0.724,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4651656547342995,
+      "learning_rate": 1.99490288881531e-05,
+      "loss": 0.6788,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3357217179420204,
+      "learning_rate": 1.994877520211519e-05,
+      "loss": 0.7071,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4526680668886418,
+      "learning_rate": 1.9948520887959772e-05,
+      "loss": 0.6997,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4834590294261867,
+      "learning_rate": 1.9948265945702905e-05,
+      "loss": 0.6164,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.429815151164952,
+      "learning_rate": 1.994801037536068e-05,
+      "loss": 0.7065,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4708895419495984,
+      "learning_rate": 1.9947754176949238e-05,
+      "loss": 0.7069,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3637496449638764,
+      "learning_rate": 1.9947497350484748e-05,
+      "loss": 0.6075,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.487163694816019,
+      "learning_rate": 1.994723989598343e-05,
+      "loss": 0.6359,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.314120518404876,
+      "learning_rate": 1.9946981813461534e-05,
+      "loss": 0.6578,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4543325677508736,
+      "learning_rate": 1.994672310293536e-05,
+      "loss": 0.7004,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4491741656825412,
+      "learning_rate": 1.9946463764421236e-05,
+      "loss": 0.7365,
+      "step": 799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5457196014736971,
+      "learning_rate": 1.994620379793554e-05,
+      "loss": 0.7442,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5056932013786395,
+      "learning_rate": 1.9945943203494677e-05,
+      "loss": 0.6771,
+      "step": 801
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4700297198243129,
+      "learning_rate": 1.994568198111511e-05,
+      "loss": 0.5979,
+      "step": 802
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.419040999628735,
+      "learning_rate": 1.9945420130813327e-05,
+      "loss": 0.7499,
+      "step": 803
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4389450439499267,
+      "learning_rate": 1.9945157652605854e-05,
+      "loss": 0.6901,
+      "step": 804
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.480339024479065,
+      "learning_rate": 1.9944894546509276e-05,
+      "loss": 0.7266,
+      "step": 805
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4854144823540691,
+      "learning_rate": 1.9944630812540188e-05,
+      "loss": 0.6809,
+      "step": 806
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.62295579712116,
+      "learning_rate": 1.9944366450715256e-05,
+      "loss": 0.695,
+      "step": 807
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3998409453367233,
+      "learning_rate": 1.994410146105116e-05,
+      "loss": 0.6088,
+      "step": 808
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3946552381773603,
+      "learning_rate": 1.9943835843564635e-05,
+      "loss": 0.6593,
+      "step": 809
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3762848226522915,
+      "learning_rate": 1.994356959827245e-05,
+      "loss": 0.7106,
+      "step": 810
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3314331591335313,
+      "learning_rate": 1.9943302725191416e-05,
+      "loss": 0.6667,
+      "step": 811
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.543037432337072,
+      "learning_rate": 1.9943035224338375e-05,
+      "loss": 0.7705,
+      "step": 812
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.315013884917083,
+      "learning_rate": 1.9942767095730225e-05,
+      "loss": 0.6175,
+      "step": 813
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.404428489151962,
+      "learning_rate": 1.994249833938389e-05,
+      "loss": 0.69,
+      "step": 814
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3887899914605013,
+      "learning_rate": 1.9942228955316342e-05,
+      "loss": 0.6877,
+      "step": 815
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5660911490464346,
+      "learning_rate": 1.994195894354458e-05,
+      "loss": 0.7207,
+      "step": 816
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.315405717275919,
+      "learning_rate": 1.9941688304085654e-05,
+      "loss": 0.6147,
+      "step": 817
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.606085517269158,
+      "learning_rate": 1.994141703695666e-05,
+      "loss": 0.6923,
+      "step": 818
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.495239492007301,
+      "learning_rate": 1.994114514217471e-05,
+      "loss": 0.6948,
+      "step": 819
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4803970037704692,
+      "learning_rate": 1.994087261975698e-05,
+      "loss": 0.7193,
+      "step": 820
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4623073744465764,
+      "learning_rate": 1.9940599469720675e-05,
+      "loss": 0.6743,
+      "step": 821
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3700775118988122,
+      "learning_rate": 1.994032569208304e-05,
+      "loss": 0.6744,
+      "step": 822
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4744522920524408,
+      "learning_rate": 1.9940051286861357e-05,
+      "loss": 0.6349,
+      "step": 823
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3809143228133267,
+      "learning_rate": 1.993977625407295e-05,
+      "loss": 0.7152,
+      "step": 824
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3811249861954153,
+      "learning_rate": 1.993950059373519e-05,
+      "loss": 0.6776,
+      "step": 825
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.387332447348377,
+      "learning_rate": 1.9939224305865474e-05,
+      "loss": 0.6825,
+      "step": 826
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4376772895619236,
+      "learning_rate": 1.993894739048125e-05,
+      "loss": 0.7792,
+      "step": 827
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4289809185059477,
+      "learning_rate": 1.9938669847599996e-05,
+      "loss": 0.7142,
+      "step": 828
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.429826656719188,
+      "learning_rate": 1.993839167723924e-05,
+      "loss": 0.7495,
+      "step": 829
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4351355138452868,
+      "learning_rate": 1.993811287941654e-05,
+      "loss": 0.698,
+      "step": 830
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6273453000519578,
+      "learning_rate": 1.9937833454149503e-05,
+      "loss": 0.6459,
+      "step": 831
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4846628010469245,
+      "learning_rate": 1.9937553401455763e-05,
+      "loss": 0.6972,
+      "step": 832
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3958818749864583,
+      "learning_rate": 1.993727272135301e-05,
+      "loss": 0.6921,
+      "step": 833
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.493055560484105,
+      "learning_rate": 1.993699141385896e-05,
+      "loss": 0.7243,
+      "step": 834
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5428751856768164,
+      "learning_rate": 1.9936709478991368e-05,
+      "loss": 0.6991,
+      "step": 835
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.433996824721222,
+      "learning_rate": 1.9936426916768047e-05,
+      "loss": 0.6577,
+      "step": 836
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.406986721759708,
+      "learning_rate": 1.9936143727206826e-05,
+      "loss": 0.6756,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4491021862605051,
+      "learning_rate": 1.993585991032559e-05,
+      "loss": 0.6922,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3861849334966856,
+      "learning_rate": 1.9935575466142256e-05,
+      "loss": 0.6396,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4831036393419366,
+      "learning_rate": 1.993529039467478e-05,
+      "loss": 0.7079,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.262685069125964,
+      "learning_rate": 1.9935004695941164e-05,
+      "loss": 0.6484,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4497742937620992,
+      "learning_rate": 1.9934718369959438e-05,
+      "loss": 0.7075,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.375661474123054,
+      "learning_rate": 1.993443141674769e-05,
+      "loss": 0.6859,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5053634915575964,
+      "learning_rate": 1.9934143836324032e-05,
+      "loss": 0.6376,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4212393545055704,
+      "learning_rate": 1.9933855628706616e-05,
+      "loss": 0.7012,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4160093951910373,
+      "learning_rate": 1.9933566793913646e-05,
+      "loss": 0.6748,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.441844015660788,
+      "learning_rate": 1.9933277331963354e-05,
+      "loss": 0.7426,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4391726009568306,
+      "learning_rate": 1.9932987242874014e-05,
+      "loss": 0.6829,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4962067324852635,
+      "learning_rate": 1.993269652666394e-05,
+      "loss": 0.6847,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5302027799827125,
+      "learning_rate": 1.9932405183351492e-05,
+      "loss": 0.7161,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3747106594485758,
+      "learning_rate": 1.993211321295506e-05,
+      "loss": 0.6486,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3602210295105055,
+      "learning_rate": 1.9931820615493075e-05,
+      "loss": 0.706,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5689407354264588,
+      "learning_rate": 1.9931527390984016e-05,
+      "loss": 0.7051,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4528060891282195,
+      "learning_rate": 1.993123353944639e-05,
+      "loss": 0.7376,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.585080897315511,
+      "learning_rate": 1.9930939060898754e-05,
+      "loss": 0.6934,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3884297735294173,
+      "learning_rate": 1.99306439553597e-05,
+      "loss": 0.6519,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.270608063832158,
+      "learning_rate": 1.9930348222847856e-05,
+      "loss": 0.6779,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2991058740998236,
+      "learning_rate": 1.9930051863381893e-05,
+      "loss": 0.6505,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3009620976098495,
+      "learning_rate": 1.9929754876980523e-05,
+      "loss": 0.5922,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3634349222659414,
+      "learning_rate": 1.99294572636625e-05,
+      "loss": 0.6267,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3665605688005118,
+      "learning_rate": 1.992915902344661e-05,
+      "loss": 0.6439,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4513939832550313,
+      "learning_rate": 1.9928860156351683e-05,
+      "loss": 0.7101,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3809437597840335,
+      "learning_rate": 1.9928560662396585e-05,
+      "loss": 0.6228,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5473122797487913,
+      "learning_rate": 1.9928260541600233e-05,
+      "loss": 0.6886,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5049679662140467,
+      "learning_rate": 1.9927959793981567e-05,
+      "loss": 0.7393,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4529344167118305,
+      "learning_rate": 1.9927658419559577e-05,
+      "loss": 0.6844,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.490682991867091,
+      "learning_rate": 1.992735641835329e-05,
+      "loss": 0.7413,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5769126692244164,
+      "learning_rate": 1.9927053790381777e-05,
+      "loss": 0.7244,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6577558329799882,
+      "learning_rate": 1.9926750535664138e-05,
+      "loss": 0.6856,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.474557554837615,
+      "learning_rate": 1.9926446654219526e-05,
+      "loss": 0.6884,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3771862941846595,
+      "learning_rate": 1.9926142146067122e-05,
+      "loss": 0.6543,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4503339810333056,
+      "learning_rate": 1.9925837011226154e-05,
+      "loss": 0.6353,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4502429890100663,
+      "learning_rate": 1.9925531249715883e-05,
+      "loss": 0.7419,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4198792253094388,
+      "learning_rate": 1.9925224861555614e-05,
+      "loss": 0.694,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.459053742624924,
+      "learning_rate": 1.992491784676469e-05,
+      "loss": 0.6742,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4462206487553946,
+      "learning_rate": 1.99246102053625e-05,
+      "loss": 0.6917,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4885586857962874,
+      "learning_rate": 1.9924301937368463e-05,
+      "loss": 0.7073,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.514222981202289,
+      "learning_rate": 1.992399304280204e-05,
+      "loss": 0.7174,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.572562129335279,
+      "learning_rate": 1.992368352168274e-05,
+      "loss": 0.7322,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4745071880984189,
+      "learning_rate": 1.9923373374030098e-05,
+      "loss": 0.6473,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4927841511558657,
+      "learning_rate": 1.9923062599863694e-05,
+      "loss": 0.7284,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4004572768689971,
+      "learning_rate": 1.9922751199203153e-05,
+      "loss": 0.694,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3811435865604569,
+      "learning_rate": 1.9922439172068134e-05,
+      "loss": 0.6778,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6417994337404593,
+      "learning_rate": 1.9922126518478338e-05,
+      "loss": 0.7517,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3503645510662545,
+      "learning_rate": 1.9921813238453505e-05,
+      "loss": 0.6337,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4140744287820188,
+      "learning_rate": 1.992149933201341e-05,
+      "loss": 0.6708,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3250569079432117,
+      "learning_rate": 1.9921184799177874e-05,
+      "loss": 0.5896,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6496578237440829,
+      "learning_rate": 1.9920869639966754e-05,
+      "loss": 0.7237,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.530620211885439,
+      "learning_rate": 1.9920553854399954e-05,
+      "loss": 0.6639,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3179543106020766,
+      "learning_rate": 1.9920237442497402e-05,
+      "loss": 0.6178,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.386334433155643,
+      "learning_rate": 1.9919920404279078e-05,
+      "loss": 0.6786,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4601150890608667,
+      "learning_rate": 1.9919602739765e-05,
+      "loss": 0.679,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4218650230644563,
+      "learning_rate": 1.991928444897522e-05,
+      "loss": 0.6331,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5476835045299449,
+      "learning_rate": 1.991896553192984e-05,
+      "loss": 0.663,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3503831778782487,
+      "learning_rate": 1.9918645988648988e-05,
+      "loss": 0.6662,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.488689696966759,
+      "learning_rate": 1.991832581915284e-05,
+      "loss": 0.7097,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5004022376840858,
+      "learning_rate": 1.991800502346162e-05,
+      "loss": 0.6784,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5633060664962397,
+      "learning_rate": 1.9917683601595563e-05,
+      "loss": 0.7051,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4859099633265669,
+      "learning_rate": 1.9917361553574974e-05,
+      "loss": 0.7132,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2765667619949221,
+      "learning_rate": 1.9917038879420184e-05,
+      "loss": 0.6212,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4555703373919575,
+      "learning_rate": 1.9916715579151567e-05,
+      "loss": 0.7504,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4165845267474468,
+      "learning_rate": 1.9916391652789526e-05,
+      "loss": 0.659,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4645597055337136,
+      "learning_rate": 1.9916067100354523e-05,
+      "loss": 0.7273,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3920263023686752,
+      "learning_rate": 1.9915741921867043e-05,
+      "loss": 0.6225,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4166586632595999,
+      "learning_rate": 1.9915416117347615e-05,
+      "loss": 0.6823,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3831162630917924,
+      "learning_rate": 1.9915089686816813e-05,
+      "loss": 0.7372,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4050313119351863,
+      "learning_rate": 1.991476263029524e-05,
+      "loss": 0.6783,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4914840717903826,
+      "learning_rate": 1.9914434947803552e-05,
+      "loss": 0.6823,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.487423066145206,
+      "learning_rate": 1.9914106639362437e-05,
+      "loss": 0.7083,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5068998589001918,
+      "learning_rate": 1.9913777704992613e-05,
+      "loss": 0.7593,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3678746267863946,
+      "learning_rate": 1.991344814471486e-05,
+      "loss": 0.6488,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3588297397128217,
+      "learning_rate": 1.9913117958549975e-05,
+      "loss": 0.651,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.40965020988121,
+      "learning_rate": 1.9912787146518812e-05,
+      "loss": 0.6457,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4230327713356739,
+      "learning_rate": 1.991245570864225e-05,
+      "loss": 0.6117,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3945193250130232,
+      "learning_rate": 1.9912123644941218e-05,
+      "loss": 0.6508,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.493799826457363,
+      "learning_rate": 1.9911790955436682e-05,
+      "loss": 0.7071,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4551092566929722,
+      "learning_rate": 1.9911457640149642e-05,
+      "loss": 0.6186,
+      "step": 917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4483474621158612,
+      "learning_rate": 1.991112369910115e-05,
+      "loss": 0.652,
+      "step": 918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3635141780327364,
+      "learning_rate": 1.9910789132312278e-05,
+      "loss": 0.6612,
+      "step": 919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4896489464596532,
+      "learning_rate": 1.9910453939804156e-05,
+      "loss": 0.6837,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4126587955708032,
+      "learning_rate": 1.991011812159795e-05,
+      "loss": 0.6654,
+      "step": 921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8112980375902783,
+      "learning_rate": 1.9909781677714854e-05,
+      "loss": 0.6903,
+      "step": 922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.541463383168267,
+      "learning_rate": 1.9909444608176117e-05,
+      "loss": 0.6427,
+      "step": 923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3863376577333875,
+      "learning_rate": 1.9909106913003013e-05,
+      "loss": 0.6712,
+      "step": 924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5331110641150918,
+      "learning_rate": 1.9908768592216862e-05,
+      "loss": 0.7042,
+      "step": 925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4753000827000902,
+      "learning_rate": 1.990842964583903e-05,
+      "loss": 0.6313,
+      "step": 926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.440826382080631,
+      "learning_rate": 1.9908090073890915e-05,
+      "loss": 0.7069,
+      "step": 927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3562332328076356,
+      "learning_rate": 1.9907749876393955e-05,
+      "loss": 0.654,
+      "step": 928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4162225401085708,
+      "learning_rate": 1.9907409053369625e-05,
+      "loss": 0.6433,
+      "step": 929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3852365228465156,
+      "learning_rate": 1.990706760483945e-05,
+      "loss": 0.6739,
+      "step": 930
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4268325900953065,
+      "learning_rate": 1.990672553082498e-05,
+      "loss": 0.7006,
+      "step": 931
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4229980895993992,
+      "learning_rate": 1.9906382831347815e-05,
+      "loss": 0.7259,
+      "step": 932
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.429748533853472,
+      "learning_rate": 1.9906039506429594e-05,
+      "loss": 0.7383,
+      "step": 933
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3603332036320621,
+      "learning_rate": 1.990569555609199e-05,
+      "loss": 0.6462,
+      "step": 934
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5254504867315368,
+      "learning_rate": 1.9905350980356717e-05,
+      "loss": 0.7234,
+      "step": 935
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3442592543002676,
+      "learning_rate": 1.990500577924553e-05,
+      "loss": 0.6751,
+      "step": 936
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3598834435219844,
+      "learning_rate": 1.990465995278023e-05,
+      "loss": 0.7251,
+      "step": 937
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3867053823767008,
+      "learning_rate": 1.9904313500982645e-05,
+      "loss": 0.7055,
+      "step": 938
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2992096992952753,
+      "learning_rate": 1.9903966423874648e-05,
+      "loss": 0.6868,
+      "step": 939
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4236460963139852,
+      "learning_rate": 1.9903618721478154e-05,
+      "loss": 0.6872,
+      "step": 940
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3349318706972957,
+      "learning_rate": 1.9903270393815112e-05,
+      "loss": 0.6536,
+      "step": 941
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5064617688746478,
+      "learning_rate": 1.9902921440907517e-05,
+      "loss": 0.7324,
+      "step": 942
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4943537940256149,
+      "learning_rate": 1.99025718627774e-05,
+      "loss": 0.7533,
+      "step": 943
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.326285490806938,
+      "learning_rate": 1.990222165944683e-05,
+      "loss": 0.6096,
+      "step": 944
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4308625275631561,
+      "learning_rate": 1.9901870830937916e-05,
+      "loss": 0.7198,
+      "step": 945
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4860911036221065,
+      "learning_rate": 1.9901519377272813e-05,
+      "loss": 0.7275,
+      "step": 946
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3758716421418185,
+      "learning_rate": 1.9901167298473704e-05,
+      "loss": 0.6426,
+      "step": 947
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3939826377734255,
+      "learning_rate": 1.990081459456282e-05,
+      "loss": 0.6446,
+      "step": 948
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4762685170251857,
+      "learning_rate": 1.990046126556243e-05,
+      "loss": 0.6721,
+      "step": 949
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.316357122121463,
+      "learning_rate": 1.9900107311494844e-05,
+      "loss": 0.6468,
+      "step": 950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4429846084789009,
+      "learning_rate": 1.98997527323824e-05,
+      "loss": 0.7011,
+      "step": 951
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4064551733598605,
+      "learning_rate": 1.9899397528247496e-05,
+      "loss": 0.687,
+      "step": 952
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4627218518349863,
+      "learning_rate": 1.9899041699112547e-05,
+      "loss": 0.6897,
+      "step": 953
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3036922382056675,
+      "learning_rate": 1.9898685245000023e-05,
+      "loss": 0.6743,
+      "step": 954
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.487256195219612,
+      "learning_rate": 1.9898328165932432e-05,
+      "loss": 0.6504,
+      "step": 955
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.488864574107309,
+      "learning_rate": 1.989797046193231e-05,
+      "loss": 0.6602,
+      "step": 956
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4369606996708997,
+      "learning_rate": 1.9897612133022252e-05,
+      "loss": 0.6301,
+      "step": 957
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3805988934778417,
+      "learning_rate": 1.9897253179224872e-05,
+      "loss": 0.6489,
+      "step": 958
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.24645412101121,
+      "learning_rate": 1.9896893600562836e-05,
+      "loss": 0.6412,
+      "step": 959
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5830254339029253,
+      "learning_rate": 1.9896533397058846e-05,
+      "loss": 0.7216,
+      "step": 960
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4225778205113302,
+      "learning_rate": 1.9896172568735642e-05,
+      "loss": 0.6761,
+      "step": 961
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.425753825720842,
+      "learning_rate": 1.989581111561601e-05,
+      "loss": 0.6642,
+      "step": 962
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3112592736453563,
+      "learning_rate": 1.989544903772276e-05,
+      "loss": 0.6485,
+      "step": 963
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3623783652231403,
+      "learning_rate": 1.9895086335078767e-05,
+      "loss": 0.6011,
+      "step": 964
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.446612129472585,
+      "learning_rate": 1.9894723007706916e-05,
+      "loss": 0.7153,
+      "step": 965
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.330918549301755,
+      "learning_rate": 1.989435905563015e-05,
+      "loss": 0.679,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3840427421904582,
+      "learning_rate": 1.9893994478871456e-05,
+      "loss": 0.6383,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3659906434678413,
+      "learning_rate": 1.989362927745384e-05,
+      "loss": 0.7313,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4293000576987829,
+      "learning_rate": 1.989326345140036e-05,
+      "loss": 0.6224,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3549320796539817,
+      "learning_rate": 1.9892897000734122e-05,
+      "loss": 0.6781,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7309566869049375,
+      "learning_rate": 1.989252992547825e-05,
+      "loss": 0.5915,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.471031932372707,
+      "learning_rate": 1.989216222565593e-05,
+      "loss": 0.6853,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.29330896240718,
+      "learning_rate": 1.9891793901290367e-05,
+      "loss": 0.5932,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.39955222598446,
+      "learning_rate": 1.9891424952404825e-05,
+      "loss": 0.6759,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3091365314821897,
+      "learning_rate": 1.989105537902259e-05,
+      "loss": 0.6615,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3550494862067861,
+      "learning_rate": 1.9890685181166996e-05,
+      "loss": 0.6901,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4487513690774159,
+      "learning_rate": 1.989031435886142e-05,
+      "loss": 0.6796,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5008200152433027,
+      "learning_rate": 1.988994291212927e-05,
+      "loss": 0.6849,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.474959214503317,
+      "learning_rate": 1.9889570840994e-05,
+      "loss": 0.6223,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4913596208415856,
+      "learning_rate": 1.98891981454791e-05,
+      "loss": 0.7232,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2386380232194563,
+      "learning_rate": 1.9888824825608097e-05,
+      "loss": 0.6471,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5064282165451475,
+      "learning_rate": 1.9888450881404563e-05,
+      "loss": 0.7048,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4884845264413473,
+      "learning_rate": 1.988807631289211e-05,
+      "loss": 0.7184,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3160290729563668,
+      "learning_rate": 1.988770112009438e-05,
+      "loss": 0.6145,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4277086581772478,
+      "learning_rate": 1.9887325303035063e-05,
+      "loss": 0.7313,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3675087687544234,
+      "learning_rate": 1.9886948861737892e-05,
+      "loss": 0.6608,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.531677342164226,
+      "learning_rate": 1.988657179622663e-05,
+      "loss": 0.7273,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4073944097587059,
+      "learning_rate": 1.9886194106525082e-05,
+      "loss": 0.7083,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3219467522376414,
+      "learning_rate": 1.988581579265709e-05,
+      "loss": 0.6774,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4440426583527428,
+      "learning_rate": 1.9885436854646546e-05,
+      "loss": 0.6863,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.439198527083775,
+      "learning_rate": 1.9885057292517373e-05,
+      "loss": 0.6843,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2199872901129378,
+      "learning_rate": 1.9884677106293528e-05,
+      "loss": 0.643,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3182833275230958,
+      "learning_rate": 1.9884296295999022e-05,
+      "loss": 0.6841,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3048183809522353,
+      "learning_rate": 1.9883914861657895e-05,
+      "loss": 0.6242,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4444066688700568,
+      "learning_rate": 1.988353280329423e-05,
+      "loss": 0.7121,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.382637012790591,
+      "learning_rate": 1.9883150120932144e-05,
+      "loss": 0.6451,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3578731966664257,
+      "learning_rate": 1.98827668145958e-05,
+      "loss": 0.6442,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3944573476320203,
+      "learning_rate": 1.9882382884309402e-05,
+      "loss": 0.7148,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3060491161019383,
+      "learning_rate": 1.9881998330097184e-05,
+      "loss": 0.6531,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4090578657062638,
+      "learning_rate": 1.9881613151983425e-05,
+      "loss": 0.7014,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3586187341727907,
+      "learning_rate": 1.9881227349992448e-05,
+      "loss": 0.6605,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.459980044032585,
+      "learning_rate": 1.9880840924148606e-05,
+      "loss": 0.7058,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2951328571759797,
+      "learning_rate": 1.9880453874476302e-05,
+      "loss": 0.6291,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3904379815348435,
+      "learning_rate": 1.9880066200999963e-05,
+      "loss": 0.6767,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5176683789372114,
+      "learning_rate": 1.9879677903744076e-05,
+      "loss": 0.7051,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4120921955672878,
+      "learning_rate": 1.9879288982733146e-05,
+      "loss": 0.6749,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3639943790795412,
+      "learning_rate": 1.9878899437991736e-05,
+      "loss": 0.6627,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.333326945686934,
+      "learning_rate": 1.987850926954443e-05,
+      "loss": 0.6463,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.433813426469793,
+      "learning_rate": 1.987811847741587e-05,
+      "loss": 0.6474,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4291352419646783,
+      "learning_rate": 1.9877727061630732e-05,
+      "loss": 0.6823,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.318844808994173,
+      "learning_rate": 1.9877335022213716e-05,
+      "loss": 0.6547,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3555024772207713,
+      "learning_rate": 1.987694235918958e-05,
+      "loss": 0.6181,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3196808246332612,
+      "learning_rate": 1.9876549072583116e-05,
+      "loss": 0.6266,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2748886452460877,
+      "learning_rate": 1.9876155162419155e-05,
+      "loss": 0.6764,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5086678564280764,
+      "learning_rate": 1.987576062872256e-05,
+      "loss": 0.6756,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4816374581843959,
+      "learning_rate": 1.9875365471518248e-05,
+      "loss": 0.672,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3192516782835608,
+      "learning_rate": 1.987496969083116e-05,
+      "loss": 0.6507,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.453259615661739,
+      "learning_rate": 1.987457328668629e-05,
+      "loss": 0.6735,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4272838460945279,
+      "learning_rate": 1.987417625910866e-05,
+      "loss": 0.6806,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4206384165273391,
+      "learning_rate": 1.987377860812334e-05,
+      "loss": 0.7037,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6001073413843292,
+      "learning_rate": 1.9873380333755437e-05,
+      "loss": 0.684,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3913517028294793,
+      "learning_rate": 1.987298143603009e-05,
+      "loss": 0.7063,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2660852702051149,
+      "learning_rate": 1.9872581914972486e-05,
+      "loss": 0.583,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3821509982728633,
+      "learning_rate": 1.9872181770607852e-05,
+      "loss": 0.6473,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4060382683621004,
+      "learning_rate": 1.987178100296145e-05,
+      "loss": 0.6761,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4819007731641538,
+      "learning_rate": 1.987137961205858e-05,
+      "loss": 0.7451,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3173853009584309,
+      "learning_rate": 1.9870977597924586e-05,
+      "loss": 0.6535,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3595085078482159,
+      "learning_rate": 1.987057496058485e-05,
+      "loss": 0.6398,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.306198615564732,
+      "learning_rate": 1.987017170006479e-05,
+      "loss": 0.615,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4706134743488104,
+      "learning_rate": 1.9869767816389867e-05,
+      "loss": 0.702,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3656120160007177,
+      "learning_rate": 1.986936330958558e-05,
+      "loss": 0.6992,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.498139499153313,
+      "learning_rate": 1.986895817967747e-05,
+      "loss": 0.6588,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5089207662733624,
+      "learning_rate": 1.9868552426691113e-05,
+      "loss": 0.6933,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4074484063240944,
+      "learning_rate": 1.9868146050652123e-05,
+      "loss": 0.6611,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4470945352387474,
+      "learning_rate": 1.9867739051586163e-05,
+      "loss": 0.6358,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.29824660053539,
+      "learning_rate": 1.9867331429518922e-05,
+      "loss": 0.6703,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3373726614826371,
+      "learning_rate": 1.9866923184476143e-05,
+      "loss": 0.652,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6000545194397107,
+      "learning_rate": 1.9866514316483597e-05,
+      "loss": 0.7112,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4534900985973662,
+      "learning_rate": 1.9866104825567096e-05,
+      "loss": 0.6566,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4244360862043022,
+      "learning_rate": 1.9865694711752498e-05,
+      "loss": 0.6325,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4065804411161942,
+      "learning_rate": 1.986528397506569e-05,
+      "loss": 0.6604,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2632019489902635,
+      "learning_rate": 1.9864872615532605e-05,
+      "loss": 0.6187,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.401284585571879,
+      "learning_rate": 1.9864460633179215e-05,
+      "loss": 0.6613,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6324542807096245,
+      "learning_rate": 1.9864048028031535e-05,
+      "loss": 0.6717,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4310049855100742,
+      "learning_rate": 1.9863634800115606e-05,
+      "loss": 0.6598,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.420662667026436,
+      "learning_rate": 1.9863220949457528e-05,
+      "loss": 0.6392,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.423003283536424,
+      "learning_rate": 1.9862806476083422e-05,
+      "loss": 0.6136,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3952984662861372,
+      "learning_rate": 1.9862391380019458e-05,
+      "loss": 0.7323,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4141107624732887,
+      "learning_rate": 1.9861975661291845e-05,
+      "loss": 0.6745,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3056460759501103,
+      "learning_rate": 1.9861559319926825e-05,
+      "loss": 0.6908,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.309622289207852,
+      "learning_rate": 1.9861142355950685e-05,
+      "loss": 0.6436,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.333428642720928,
+      "learning_rate": 1.9860724769389754e-05,
+      "loss": 0.6094,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4631047620518394,
+      "learning_rate": 1.986030656027039e-05,
+      "loss": 0.6917,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3177733524650974,
+      "learning_rate": 1.9859887728619002e-05,
+      "loss": 0.6302,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.44428839614962,
+      "learning_rate": 1.9859468274462034e-05,
+      "loss": 0.647,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.386861702441382,
+      "learning_rate": 1.9859048197825963e-05,
+      "loss": 0.6952,
+      "step": 1056
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4660066185056326,
+      "learning_rate": 1.9858627498737313e-05,
+      "loss": 0.7264,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3794503151745687,
+      "learning_rate": 1.9858206177222646e-05,
+      "loss": 0.5743,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.447135888589788,
+      "learning_rate": 1.9857784233308562e-05,
+      "loss": 0.6745,
+      "step": 1059
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3665941967396986,
+      "learning_rate": 1.9857361667021696e-05,
+      "loss": 0.5913,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3165333400696055,
+      "learning_rate": 1.9856938478388735e-05,
+      "loss": 0.6582,
+      "step": 1061
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3262803675188708,
+      "learning_rate": 1.9856514667436393e-05,
+      "loss": 0.6925,
+      "step": 1062
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5519308062200847,
+      "learning_rate": 1.9856090234191424e-05,
+      "loss": 0.6948,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5297323218413756,
+      "learning_rate": 1.985566517868063e-05,
+      "loss": 0.6706,
+      "step": 1064
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3614990915133625,
+      "learning_rate": 1.9855239500930846e-05,
+      "loss": 0.7215,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.426998340089791,
+      "learning_rate": 1.9854813200968942e-05,
+      "loss": 0.6469,
+      "step": 1066
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4413188297878274,
+      "learning_rate": 1.985438627882184e-05,
+      "loss": 0.6302,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2690913450583652,
+      "learning_rate": 1.9853958734516487e-05,
+      "loss": 0.5659,
+      "step": 1068
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5116095293988812,
+      "learning_rate": 1.9853530568079882e-05,
+      "loss": 0.7077,
+      "step": 1069
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3332191607150499,
+      "learning_rate": 1.9853101779539052e-05,
+      "loss": 0.6324,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3016776813589728,
+      "learning_rate": 1.9852672368921074e-05,
+      "loss": 0.6219,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4503435977428494,
+      "learning_rate": 1.9852242336253057e-05,
+      "loss": 0.6758,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4257728889951766,
+      "learning_rate": 1.985181168156215e-05,
+      "loss": 0.6645,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3685159822840158,
+      "learning_rate": 1.985138040487554e-05,
+      "loss": 0.6428,
+      "step": 1074
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4001830730629623,
+      "learning_rate": 1.985094850622046e-05,
+      "loss": 0.6404,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5185190472606231,
+      "learning_rate": 1.985051598562418e-05,
+      "loss": 0.6866,
+      "step": 1076
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4118465110332843,
+      "learning_rate": 1.9850082843114e-05,
+      "loss": 0.621,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4080094245248436,
+      "learning_rate": 1.984964907871727e-05,
+      "loss": 0.7097,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.454160669891905,
+      "learning_rate": 1.9849214692461375e-05,
+      "loss": 0.6783,
+      "step": 1079
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3085695634571022,
+      "learning_rate": 1.9848779684373746e-05,
+      "loss": 0.671,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4117438765349613,
+      "learning_rate": 1.9848344054481838e-05,
+      "loss": 0.6425,
+      "step": 1081
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4725912469334606,
+      "learning_rate": 1.9847907802813165e-05,
+      "loss": 0.672,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4635007678343672,
+      "learning_rate": 1.9847470929395257e-05,
+      "loss": 0.7695,
+      "step": 1083
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4571029195468161,
+      "learning_rate": 1.9847033434255708e-05,
+      "loss": 0.7051,
+      "step": 1084
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.309554473303262,
+      "learning_rate": 1.984659531742213e-05,
+      "loss": 0.6515,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2763513098548585,
+      "learning_rate": 1.984615657892219e-05,
+      "loss": 0.6556,
+      "step": 1086
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3691668324268231,
+      "learning_rate": 1.9845717218783588e-05,
+      "loss": 0.6564,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2869796571791097,
+      "learning_rate": 1.9845277237034057e-05,
+      "loss": 0.6352,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3807977341594648,
+      "learning_rate": 1.9844836633701383e-05,
+      "loss": 0.6988,
+      "step": 1089
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3444731341252991,
+      "learning_rate": 1.9844395408813376e-05,
+      "loss": 0.6743,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.307205420254287,
+      "learning_rate": 1.9843953562397896e-05,
+      "loss": 0.6507,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.384103764814393,
+      "learning_rate": 1.984351109448284e-05,
+      "loss": 0.7041,
+      "step": 1092
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.311451084075518,
+      "learning_rate": 1.9843068005096146e-05,
+      "loss": 0.6769,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.475281093774608,
+      "learning_rate": 1.9842624294265783e-05,
+      "loss": 0.6757,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2847458204330213,
+      "learning_rate": 1.9842179962019768e-05,
+      "loss": 0.629,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.477745510687581,
+      "learning_rate": 1.9841735008386154e-05,
+      "loss": 0.7262,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.419998775535378,
+      "learning_rate": 1.984128943339303e-05,
+      "loss": 0.6711,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3256467698183303,
+      "learning_rate": 1.984084323706853e-05,
+      "loss": 0.6787,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4531825884820975,
+      "learning_rate": 1.9840396419440825e-05,
+      "loss": 0.6695,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3048689481914824,
+      "learning_rate": 1.9839948980538124e-05,
+      "loss": 0.6897,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4104249650644398,
+      "learning_rate": 1.9839500920388676e-05,
+      "loss": 0.6653,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4111668988503194,
+      "learning_rate": 1.983905223902077e-05,
+      "loss": 0.7027,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.36188376428691,
+      "learning_rate": 1.9838602936462732e-05,
+      "loss": 0.6676,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2403181873747628,
+      "learning_rate": 1.983815301274293e-05,
+      "loss": 0.6282,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3271663122818982,
+      "learning_rate": 1.9837702467889772e-05,
+      "loss": 0.6733,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4194330898476917,
+      "learning_rate": 1.9837251301931704e-05,
+      "loss": 0.6505,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2082126107692968,
+      "learning_rate": 1.9836799514897204e-05,
+      "loss": 0.6707,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3665789312408116,
+      "learning_rate": 1.98363471068148e-05,
+      "loss": 0.6748,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3399140809398349,
+      "learning_rate": 1.9835894077713055e-05,
+      "loss": 0.6376,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3799549469987997,
+      "learning_rate": 1.9835440427620568e-05,
+      "loss": 0.677,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2417224517340832,
+      "learning_rate": 1.9834986156565984e-05,
+      "loss": 0.6271,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2371500907728634,
+      "learning_rate": 1.9834531264577984e-05,
+      "loss": 0.6623,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.415355533780316,
+      "learning_rate": 1.9834075751685283e-05,
+      "loss": 0.6715,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.555496312141032,
+      "learning_rate": 1.9833619617916645e-05,
+      "loss": 0.6729,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.52222421700903,
+      "learning_rate": 1.9833162863300863e-05,
+      "loss": 0.6321,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4989224378128634,
+      "learning_rate": 1.983270548786678e-05,
+      "loss": 0.7222,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4995699901928754,
+      "learning_rate": 1.9832247491643267e-05,
+      "loss": 0.7012,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.333947780213303,
+      "learning_rate": 1.983178887465924e-05,
+      "loss": 0.6509,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.647961588495242,
+      "learning_rate": 1.983132963694366e-05,
+      "loss": 0.8132,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5941082420883146,
+      "learning_rate": 1.9830869778525517e-05,
+      "loss": 0.6881,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4562004932289037,
+      "learning_rate": 1.9830409299433845e-05,
+      "loss": 0.6922,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3253266867481204,
+      "learning_rate": 1.9829948199697713e-05,
+      "loss": 0.6131,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3194267868511274,
+      "learning_rate": 1.9829486479346237e-05,
+      "loss": 0.6601,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3032366057614475,
+      "learning_rate": 1.9829024138408567e-05,
+      "loss": 0.6463,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3526579105754335,
+      "learning_rate": 1.982856117691389e-05,
+      "loss": 0.6676,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4044047647130808,
+      "learning_rate": 1.9828097594891435e-05,
+      "loss": 0.6607,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4112826680270278,
+      "learning_rate": 1.982763339237048e-05,
+      "loss": 0.5893,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4209828354982046,
+      "learning_rate": 1.982716856938032e-05,
+      "loss": 0.6859,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3437338983325133,
+      "learning_rate": 1.9826703125950302e-05,
+      "loss": 0.6214,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3298890459774835,
+      "learning_rate": 1.982623706210982e-05,
+      "loss": 0.6727,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3315551443910691,
+      "learning_rate": 1.98257703778883e-05,
+      "loss": 0.6581,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2907969972106836,
+      "learning_rate": 1.9825303073315197e-05,
+      "loss": 0.6066,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2909279936396814,
+      "learning_rate": 1.9824835148420023e-05,
+      "loss": 0.5687,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.424877894339963,
+      "learning_rate": 1.982436660323231e-05,
+      "loss": 0.7123,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4897768210633964,
+      "learning_rate": 1.9823897437781654e-05,
+      "loss": 0.7402,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4471998109666606,
+      "learning_rate": 1.9823427652097663e-05,
+      "loss": 0.6692,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4191902721067329,
+      "learning_rate": 1.9822957246210003e-05,
+      "loss": 0.6479,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4409011741597924,
+      "learning_rate": 1.9822486220148376e-05,
+      "loss": 0.6955,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3597189369504,
+      "learning_rate": 1.9822014573942514e-05,
+      "loss": 0.6567,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4575559269691574,
+      "learning_rate": 1.98215423076222e-05,
+      "loss": 0.6694,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3492958599027343,
+      "learning_rate": 1.982106942121724e-05,
+      "loss": 0.6675,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3921489723089153,
+      "learning_rate": 1.9820595914757505e-05,
+      "loss": 0.649,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3440147848735216,
+      "learning_rate": 1.9820121788272873e-05,
+      "loss": 0.6429,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3360572002837734,
+      "learning_rate": 1.9819647041793298e-05,
+      "loss": 0.7376,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2438325844955456,
+      "learning_rate": 1.9819171675348736e-05,
+      "loss": 0.6687,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3711687377246966,
+      "learning_rate": 1.9818695688969207e-05,
+      "loss": 0.7174,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3686320929061715,
+      "learning_rate": 1.9818219082684762e-05,
+      "loss": 0.6525,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4988193633991862,
+      "learning_rate": 1.9817741856525494e-05,
+      "loss": 0.6038,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.430496903897148,
+      "learning_rate": 1.9817264010521524e-05,
+      "loss": 0.678,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3119959998883932,
+      "learning_rate": 1.981678554470303e-05,
+      "loss": 0.6368,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5116578713315048,
+      "learning_rate": 1.9816306459100218e-05,
+      "loss": 0.7125,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3733899486933328,
+      "learning_rate": 1.981582675374333e-05,
+      "loss": 0.7114,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4422723912540691,
+      "learning_rate": 1.9815346428662655e-05,
+      "loss": 0.6551,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2559756022015318,
+      "learning_rate": 1.9814865483888527e-05,
+      "loss": 0.6528,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.374362104007575,
+      "learning_rate": 1.9814383919451298e-05,
+      "loss": 0.6326,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.340186606267172,
+      "learning_rate": 1.9813901735381373e-05,
+      "loss": 0.6691,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4115398097246712,
+      "learning_rate": 1.9813418931709203e-05,
+      "loss": 0.6561,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.382567173807268,
+      "learning_rate": 1.9812935508465263e-05,
+      "loss": 0.6679,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4407357824929055,
+      "learning_rate": 1.981245146568008e-05,
+      "loss": 0.694,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.187108276406847,
+      "learning_rate": 1.9811966803384208e-05,
+      "loss": 0.6749,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3297259835076982,
+      "learning_rate": 1.9811481521608245e-05,
+      "loss": 0.6108,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.335911756122454,
+      "learning_rate": 1.981099562038284e-05,
+      "loss": 0.6605,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3371802455832802,
+      "learning_rate": 1.981050909973866e-05,
+      "loss": 0.6176,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3062167314007176,
+      "learning_rate": 1.981002195970642e-05,
+      "loss": 0.5866,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.32331668883329,
+      "learning_rate": 1.980953420031689e-05,
+      "loss": 0.6782,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.202468742630543,
+      "learning_rate": 1.980904582160085e-05,
+      "loss": 0.5988,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2650604401726164,
+      "learning_rate": 1.980855682358914e-05,
+      "loss": 0.6952,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3592897585747172,
+      "learning_rate": 1.9808067206312632e-05,
+      "loss": 0.6597,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4574664491223743,
+      "learning_rate": 1.980757696980224e-05,
+      "loss": 0.5817,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4828517124791651,
+      "learning_rate": 1.980708611408891e-05,
+      "loss": 0.739,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2886534995682364,
+      "learning_rate": 1.9806594639203637e-05,
+      "loss": 0.6297,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5097725733588492,
+      "learning_rate": 1.980610254517745e-05,
+      "loss": 0.7327,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3970631500448782,
+      "learning_rate": 1.980560983204142e-05,
+      "loss": 0.7241,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.215500730620227,
+      "learning_rate": 1.9805116499826646e-05,
+      "loss": 0.6264,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.360247978434702,
+      "learning_rate": 1.980462254856428e-05,
+      "loss": 0.6927,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2213126407984554,
+      "learning_rate": 1.9804127978285506e-05,
+      "loss": 0.6612,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4275996905255162,
+      "learning_rate": 1.9803632789021553e-05,
+      "loss": 0.7202,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2579692482275089,
+      "learning_rate": 1.9803136980803687e-05,
+      "loss": 0.6404,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2835999779483394,
+      "learning_rate": 1.9802640553663196e-05,
+      "loss": 0.6448,
+      "step": 1179
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3310670915055065,
+      "learning_rate": 1.980214350763144e-05,
+      "loss": 0.696,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.346129904785629,
+      "learning_rate": 1.9801645842739785e-05,
+      "loss": 0.7355,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.406778321098541,
+      "learning_rate": 1.980114755901966e-05,
+      "loss": 0.7006,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3086238117465867,
+      "learning_rate": 1.9800648656502524e-05,
+      "loss": 0.6399,
+      "step": 1183
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5989505962304749,
+      "learning_rate": 1.9800149135219873e-05,
+      "loss": 0.6373,
+      "step": 1184
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4296688224531342,
+      "learning_rate": 1.9799648995203246e-05,
+      "loss": 0.6997,
+      "step": 1185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.382837887978123,
+      "learning_rate": 1.979914823648422e-05,
+      "loss": 0.663,
+      "step": 1186
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2721115417218372,
+      "learning_rate": 1.9798646859094405e-05,
+      "loss": 0.6164,
+      "step": 1187
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3852775283966918,
+      "learning_rate": 1.979814486306546e-05,
+      "loss": 0.6401,
+      "step": 1188
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3031675885237042,
+      "learning_rate": 1.9797642248429078e-05,
+      "loss": 0.6483,
+      "step": 1189
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3489120161237507,
+      "learning_rate": 1.979713901521699e-05,
+      "loss": 0.673,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2828958221770725,
+      "learning_rate": 1.9796635163460976e-05,
+      "loss": 0.6076,
+      "step": 1191
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3223242697395923,
+      "learning_rate": 1.9796130693192835e-05,
+      "loss": 0.6755,
+      "step": 1192
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4134175637100281,
+      "learning_rate": 1.979562560444442e-05,
+      "loss": 0.6333,
+      "step": 1193
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.336011427233703,
+      "learning_rate": 1.9795119897247627e-05,
+      "loss": 0.6573,
+      "step": 1194
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4294319966790596,
+      "learning_rate": 1.9794613571634378e-05,
+      "loss": 0.6447,
+      "step": 1195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3986745026150444,
+      "learning_rate": 1.979410662763664e-05,
+      "loss": 0.7732,
+      "step": 1196
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4299542954617028,
+      "learning_rate": 1.9793599065286417e-05,
+      "loss": 0.6712,
+      "step": 1197
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3742971357811897,
+      "learning_rate": 1.9793090884615756e-05,
+      "loss": 0.665,
+      "step": 1198
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.287680376250613,
+      "learning_rate": 1.9792582085656744e-05,
+      "loss": 0.6295,
+      "step": 1199
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2653865825119586,
+      "learning_rate": 1.97920726684415e-05,
+      "loss": 0.6411,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.364137528181507,
+      "learning_rate": 1.9791562633002194e-05,
+      "loss": 0.6543,
+      "step": 1201
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4053051635469254,
+      "learning_rate": 1.9791051979371017e-05,
+      "loss": 0.7137,
+      "step": 1202
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4748010371873694,
+      "learning_rate": 1.9790540707580213e-05,
+      "loss": 0.6858,
+      "step": 1203
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3118999108452385,
+      "learning_rate": 1.9790028817662065e-05,
+      "loss": 0.6154,
+      "step": 1204
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3993442089378476,
+      "learning_rate": 1.978951630964888e-05,
+      "loss": 0.6687,
+      "step": 1205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4223187770618666,
+      "learning_rate": 1.978900318357303e-05,
+      "loss": 0.6454,
+      "step": 1206
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3651858690892422,
+      "learning_rate": 1.9788489439466902e-05,
+      "loss": 0.6481,
+      "step": 1207
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.400000226497632,
+      "learning_rate": 1.9787975077362937e-05,
+      "loss": 0.6561,
+      "step": 1208
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3360728145132874,
+      "learning_rate": 1.9787460097293603e-05,
+      "loss": 0.6034,
+      "step": 1209
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.323599926777985,
+      "learning_rate": 1.9786944499291417e-05,
+      "loss": 0.6412,
+      "step": 1210
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4149072979326822,
+      "learning_rate": 1.9786428283388927e-05,
+      "loss": 0.688,
+      "step": 1211
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.414805685928908,
+      "learning_rate": 1.9785911449618732e-05,
+      "loss": 0.7068,
+      "step": 1212
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4754153846721507,
+      "learning_rate": 1.978539399801346e-05,
+      "loss": 0.6837,
+      "step": 1213
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3753551111292275,
+      "learning_rate": 1.9784875928605777e-05,
+      "loss": 0.6223,
+      "step": 1214
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4251710186780597,
+      "learning_rate": 1.978435724142839e-05,
+      "loss": 0.7228,
+      "step": 1215
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3832231763082845,
+      "learning_rate": 1.9783837936514053e-05,
+      "loss": 0.6726,
+      "step": 1216
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4500479986204673,
+      "learning_rate": 1.9783318013895552e-05,
+      "loss": 0.6548,
+      "step": 1217
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3005747588250538,
+      "learning_rate": 1.9782797473605708e-05,
+      "loss": 0.6567,
+      "step": 1218
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5355761393007654,
+      "learning_rate": 1.978227631567738e-05,
+      "loss": 0.7227,
+      "step": 1219
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4102631842732851,
+      "learning_rate": 1.9781754540143486e-05,
+      "loss": 0.742,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3439314520106866,
+      "learning_rate": 1.9781232147036958e-05,
+      "loss": 0.7207,
+      "step": 1221
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3944073790343496,
+      "learning_rate": 1.978070913639078e-05,
+      "loss": 0.676,
+      "step": 1222
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3960356731748198,
+      "learning_rate": 1.9780185508237976e-05,
+      "loss": 0.706,
+      "step": 1223
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.399362907885586,
+      "learning_rate": 1.9779661262611598e-05,
+      "loss": 0.6458,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4304060667545093,
+      "learning_rate": 1.9779136399544747e-05,
+      "loss": 0.7019,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4063710054682113,
+      "learning_rate": 1.9778610919070563e-05,
+      "loss": 0.6423,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3243160155817348,
+      "learning_rate": 1.977808482122222e-05,
+      "loss": 0.7011,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4000265800132696,
+      "learning_rate": 1.9777558106032936e-05,
+      "loss": 0.6679,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3050823413697417,
+      "learning_rate": 1.9777030773535962e-05,
+      "loss": 0.619,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2243187664490291,
+      "learning_rate": 1.977650282376459e-05,
+      "loss": 0.5934,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4588698989508349,
+      "learning_rate": 1.977597425675216e-05,
+      "loss": 0.6878,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.415595556630849,
+      "learning_rate": 1.9775445072532035e-05,
+      "loss": 0.6864,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4319039762205037,
+      "learning_rate": 1.9774915271137625e-05,
+      "loss": 0.7228,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2616836968149805,
+      "learning_rate": 1.9774384852602387e-05,
+      "loss": 0.6162,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4017493813399216,
+      "learning_rate": 1.97738538169598e-05,
+      "loss": 0.7778,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4409173896294583,
+      "learning_rate": 1.9773322164243394e-05,
+      "loss": 0.6846,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3741735228642569,
+      "learning_rate": 1.9772789894486743e-05,
+      "loss": 0.6303,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.343105560843702,
+      "learning_rate": 1.9772257007723442e-05,
+      "loss": 0.6178,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2393720377291753,
+      "learning_rate": 1.9771723503987133e-05,
+      "loss": 0.5986,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3828590083517824,
+      "learning_rate": 1.9771189383311507e-05,
+      "loss": 0.7265,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3367191727156558,
+      "learning_rate": 1.9770654645730286e-05,
+      "loss": 0.6923,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3043303314880903,
+      "learning_rate": 1.9770119291277223e-05,
+      "loss": 0.6321,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3399698626274916,
+      "learning_rate": 1.9769583319986125e-05,
+      "loss": 0.6622,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2371714338823765,
+      "learning_rate": 1.9769046731890825e-05,
+      "loss": 0.601,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.417887199305655,
+      "learning_rate": 1.9768509527025204e-05,
+      "loss": 0.6164,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3190168083716396,
+      "learning_rate": 1.9767971705423176e-05,
+      "loss": 0.6261,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3498563637085053,
+      "learning_rate": 1.97674332671187e-05,
+      "loss": 0.6424,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.430250296795552,
+      "learning_rate": 1.9766894212145767e-05,
+      "loss": 0.6578,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3691569502832655,
+      "learning_rate": 1.9766354540538414e-05,
+      "loss": 0.6788,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.339045161487761,
+      "learning_rate": 1.976581425233071e-05,
+      "loss": 0.6605,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2840355621802468,
+      "learning_rate": 1.976527334755677e-05,
+      "loss": 0.6267,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2673447310319617,
+      "learning_rate": 1.976473182625074e-05,
+      "loss": 0.6337,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.346546899612652,
+      "learning_rate": 1.9764189688446803e-05,
+      "loss": 0.6606,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4501943260728969,
+      "learning_rate": 1.9763646934179204e-05,
+      "loss": 0.7229,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2845667730685473,
+      "learning_rate": 1.97631035634822e-05,
+      "loss": 0.6559,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.397157648152016,
+      "learning_rate": 1.976255957639009e-05,
+      "loss": 0.6663,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3866889628045302,
+      "learning_rate": 1.976201497293723e-05,
+      "loss": 0.64,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2593359872535244,
+      "learning_rate": 1.9761469753158e-05,
+      "loss": 0.6902,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3070455930006084,
+      "learning_rate": 1.976092391708682e-05,
+      "loss": 0.7115,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2864158046760956,
+      "learning_rate": 1.9760377464758154e-05,
+      "loss": 0.6545,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4093333455969654,
+      "learning_rate": 1.9759830396206504e-05,
+      "loss": 0.7259,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2496139884021689,
+      "learning_rate": 1.9759282711466405e-05,
+      "loss": 0.6229,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4022444580969724,
+      "learning_rate": 1.9758734410572435e-05,
+      "loss": 0.7355,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3869299929195074,
+      "learning_rate": 1.9758185493559216e-05,
+      "loss": 0.6824,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.442074669818633,
+      "learning_rate": 1.97576359604614e-05,
+      "loss": 0.6732,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4624501113859736,
+      "learning_rate": 1.9757085811313686e-05,
+      "loss": 0.6482,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2907830517955654,
+      "learning_rate": 1.97565350461508e-05,
+      "loss": 0.6272,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3325046358153128,
+      "learning_rate": 1.9755983665007524e-05,
+      "loss": 0.6327,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4107812568951001,
+      "learning_rate": 1.975543166791866e-05,
+      "loss": 0.6973,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3711360479280659,
+      "learning_rate": 1.9754879054919067e-05,
+      "loss": 0.6563,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3825662253523505,
+      "learning_rate": 1.975432582604363e-05,
+      "loss": 0.6227,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3599147821121875,
+      "learning_rate": 1.9753771981327278e-05,
+      "loss": 0.655,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3522012336867513,
+      "learning_rate": 1.9753217520804976e-05,
+      "loss": 0.6367,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3559167215127172,
+      "learning_rate": 1.9752662444511738e-05,
+      "loss": 0.6678,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3938523464515515,
+      "learning_rate": 1.9752106752482596e-05,
+      "loss": 0.6849,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2887214440660997,
+      "learning_rate": 1.9751550444752642e-05,
+      "loss": 0.6558,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.506260047974055,
+      "learning_rate": 1.9750993521356997e-05,
+      "loss": 0.6281,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2844096510916296,
+      "learning_rate": 1.9750435982330823e-05,
+      "loss": 0.6187,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3041757276583863,
+      "learning_rate": 1.9749877827709323e-05,
+      "loss": 0.6785,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3923368311810609,
+      "learning_rate": 1.974931905752773e-05,
+      "loss": 0.6603,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4022437354845803,
+      "learning_rate": 1.9748759671821323e-05,
+      "loss": 0.6251,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2979961742380008,
+      "learning_rate": 1.9748199670625423e-05,
+      "loss": 0.6155,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4421464213661455,
+      "learning_rate": 1.9747639053975386e-05,
+      "loss": 0.69,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2610511548937215,
+      "learning_rate": 1.9747077821906602e-05,
+      "loss": 0.5852,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2882226693533523,
+      "learning_rate": 1.9746515974454508e-05,
+      "loss": 0.6923,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.409332753497726,
+      "learning_rate": 1.9745953511654573e-05,
+      "loss": 0.7094,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2603695390366016,
+      "learning_rate": 1.974539043354231e-05,
+      "loss": 0.5916,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4142534327297618,
+      "learning_rate": 1.9744826740153272e-05,
+      "loss": 0.7057,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.284603707400155,
+      "learning_rate": 1.9744262431523045e-05,
+      "loss": 0.6134,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2263677162962707,
+      "learning_rate": 1.9743697507687253e-05,
+      "loss": 0.6522,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3702936256004872,
+      "learning_rate": 1.974313196868157e-05,
+      "loss": 0.6236,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3339402734663026,
+      "learning_rate": 1.97425658145417e-05,
+      "loss": 0.6803,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3788075614388753,
+      "learning_rate": 1.974199904530338e-05,
+      "loss": 0.6654,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3536417408326378,
+      "learning_rate": 1.9741431661002403e-05,
+      "loss": 0.6845,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3127354456120703,
+      "learning_rate": 1.974086366167458e-05,
+      "loss": 0.6953,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4802084016990065,
+      "learning_rate": 1.9740295047355784e-05,
+      "loss": 0.6945,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2733851697033904,
+      "learning_rate": 1.9739725818081905e-05,
+      "loss": 0.6042,
+      "step": 1297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3102079541644611,
+      "learning_rate": 1.9739155973888887e-05,
+      "loss": 0.6492,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3442931297411025,
+      "learning_rate": 1.97385855148127e-05,
+      "loss": 0.6496,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.322587710754856,
+      "learning_rate": 1.9738014440889368e-05,
+      "loss": 0.6683,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4261772546429026,
+      "learning_rate": 1.9737442752154944e-05,
+      "loss": 0.6877,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4571349897092112,
+      "learning_rate": 1.973687044864552e-05,
+      "loss": 0.6411,
+      "step": 1302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2512433544062485,
+      "learning_rate": 1.9736297530397227e-05,
+      "loss": 0.6064,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3203760165674958,
+      "learning_rate": 1.973572399744624e-05,
+      "loss": 0.6353,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3100388021036906,
+      "learning_rate": 1.9735149849828767e-05,
+      "loss": 0.6889,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5577264729490756,
+      "learning_rate": 1.973457508758106e-05,
+      "loss": 0.7482,
+      "step": 1306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.399440512237503,
+      "learning_rate": 1.9733999710739398e-05,
+      "loss": 0.6147,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.429925283867604,
+      "learning_rate": 1.9733423719340118e-05,
+      "loss": 0.7181,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2958207038844536,
+      "learning_rate": 1.973284711341958e-05,
+      "loss": 0.6567,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5082136025863881,
+      "learning_rate": 1.9732269893014188e-05,
+      "loss": 0.7189,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2778157053492332,
+      "learning_rate": 1.9731692058160387e-05,
+      "loss": 0.6148,
+      "step": 1311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3436301865501994,
+      "learning_rate": 1.9731113608894656e-05,
+      "loss": 0.6477,
+      "step": 1312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4867034317967422,
+      "learning_rate": 1.9730534545253516e-05,
+      "loss": 0.6589,
+      "step": 1313
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5054852012127375,
+      "learning_rate": 1.9729954867273528e-05,
+      "loss": 0.7017,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.334658763650278,
+      "learning_rate": 1.9729374574991288e-05,
+      "loss": 0.6586,
+      "step": 1315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4171479006179815,
+      "learning_rate": 1.9728793668443437e-05,
+      "loss": 0.6929,
+      "step": 1316
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3340001028538426,
+      "learning_rate": 1.9728212147666647e-05,
+      "loss": 0.6363,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4973064716598479,
+      "learning_rate": 1.9727630012697633e-05,
+      "loss": 0.6183,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3671159562056532,
+      "learning_rate": 1.9727047263573148e-05,
+      "loss": 0.6643,
+      "step": 1319
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3051263219707756,
+      "learning_rate": 1.9726463900329985e-05,
+      "loss": 0.6885,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3832130498565052,
+      "learning_rate": 1.972587992300497e-05,
+      "loss": 0.6614,
+      "step": 1321
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.427180191799803,
+      "learning_rate": 1.972529533163498e-05,
+      "loss": 0.6415,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.402858374056527,
+      "learning_rate": 1.972471012625692e-05,
+      "loss": 0.6604,
+      "step": 1323
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3144097285596168,
+      "learning_rate": 1.9724124306907737e-05,
+      "loss": 0.7101,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4888946790883095,
+      "learning_rate": 1.9723537873624418e-05,
+      "loss": 0.7099,
+      "step": 1325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3923004858063108,
+      "learning_rate": 1.9722950826443983e-05,
+      "loss": 0.635,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3574628864926497,
+      "learning_rate": 1.97223631654035e-05,
+      "loss": 0.6596,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3104516348648454,
+      "learning_rate": 1.972177489054007e-05,
+      "loss": 0.6785,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.494565576696541,
+      "learning_rate": 1.972118600189083e-05,
+      "loss": 0.7091,
+      "step": 1329
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2904097943384203,
+      "learning_rate": 1.9720596499492965e-05,
+      "loss": 0.6412,
+      "step": 1330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2696502160304957,
+      "learning_rate": 1.9720006383383694e-05,
+      "loss": 0.6497,
+      "step": 1331
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3764578286946552,
+      "learning_rate": 1.971941565360027e-05,
+      "loss": 0.6016,
+      "step": 1332
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3056387716966713,
+      "learning_rate": 1.971882431017999e-05,
+      "loss": 0.6291,
+      "step": 1333
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3281560108827877,
+      "learning_rate": 1.9718232353160187e-05,
+      "loss": 0.6634,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3282381346144063,
+      "learning_rate": 1.971763978257824e-05,
+      "loss": 0.5961,
+      "step": 1335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3181828133361073,
+      "learning_rate": 1.9717046598471553e-05,
+      "loss": 0.6113,
+      "step": 1336
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2592238096444863,
+      "learning_rate": 1.9716452800877582e-05,
+      "loss": 0.6382,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3874714272796158,
+      "learning_rate": 1.9715858389833816e-05,
+      "loss": 0.6665,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4367131277184888,
+      "learning_rate": 1.9715263365377782e-05,
+      "loss": 0.6465,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3441409052543936,
+      "learning_rate": 1.9714667727547047e-05,
+      "loss": 0.6306,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2598321468325648,
+      "learning_rate": 1.9714071476379216e-05,
+      "loss": 0.6457,
+      "step": 1341
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2656911667847015,
+      "learning_rate": 1.971347461191194e-05,
+      "loss": 0.6455,
+      "step": 1342
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.393301114039878,
+      "learning_rate": 1.9712877134182893e-05,
+      "loss": 0.6345,
+      "step": 1343
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2886923518639681,
+      "learning_rate": 1.97122790432298e-05,
+      "loss": 0.6618,
+      "step": 1344
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3950811418568103,
+      "learning_rate": 1.9711680339090425e-05,
+      "loss": 0.7288,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4447502073441947,
+      "learning_rate": 1.9711081021802563e-05,
+      "loss": 0.7497,
+      "step": 1346
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2850127444395998,
+      "learning_rate": 1.9710481091404056e-05,
+      "loss": 0.6443,
+      "step": 1347
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4118943004462583,
+      "learning_rate": 1.9709880547932774e-05,
+      "loss": 0.6737,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3070703549643141,
+      "learning_rate": 1.9709279391426638e-05,
+      "loss": 0.649,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2349974113313673,
+      "learning_rate": 1.97086776219236e-05,
+      "loss": 0.6706,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3275052139122363,
+      "learning_rate": 1.9708075239461656e-05,
+      "loss": 0.5965,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1952545550893299,
+      "learning_rate": 1.9707472244078833e-05,
+      "loss": 0.6654,
+      "step": 1352
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2581092058034615,
+      "learning_rate": 1.9706868635813205e-05,
+      "loss": 0.6484,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3635989368244346,
+      "learning_rate": 1.970626441470288e-05,
+      "loss": 0.7099,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.352673773822253,
+      "learning_rate": 1.9705659580785997e-05,
+      "loss": 0.7159,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4107267963827326,
+      "learning_rate": 1.9705054134100758e-05,
+      "loss": 0.65,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.417609891853523,
+      "learning_rate": 1.9704448074685377e-05,
+      "loss": 0.6832,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2732380435815898,
+      "learning_rate": 1.9703841402578122e-05,
+      "loss": 0.6452,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.281590346745653,
+      "learning_rate": 1.9703234117817293e-05,
+      "loss": 0.628,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3332260903303768,
+      "learning_rate": 1.9702626220441234e-05,
+      "loss": 0.5842,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4164228603527387,
+      "learning_rate": 1.970201771048832e-05,
+      "loss": 0.6811,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.337874987434274,
+      "learning_rate": 1.9701408587996976e-05,
+      "loss": 0.6857,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3376260983215107,
+      "learning_rate": 1.9700798853005652e-05,
+      "loss": 0.6693,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.46019224037812,
+      "learning_rate": 1.9700188505552847e-05,
+      "loss": 0.6573,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3934939076482433,
+      "learning_rate": 1.9699577545677097e-05,
+      "loss": 0.6704,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3225548116650585,
+      "learning_rate": 1.969896597341697e-05,
+      "loss": 0.6682,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3609855144102638,
+      "learning_rate": 1.9698353788811083e-05,
+      "loss": 0.666,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1670668063388954,
+      "learning_rate": 1.9697740991898087e-05,
+      "loss": 0.6495,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3525111667221432,
+      "learning_rate": 1.9697127582716664e-05,
+      "loss": 0.6399,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3967121925460386,
+      "learning_rate": 1.9696513561305548e-05,
+      "loss": 0.7391,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2661433159340119,
+      "learning_rate": 1.96958989277035e-05,
+      "loss": 0.6653,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4694807791268958,
+      "learning_rate": 1.9695283681949337e-05,
+      "loss": 0.6338,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3232123677740102,
+      "learning_rate": 1.9694667824081888e-05,
+      "loss": 0.6726,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4554458460743576,
+      "learning_rate": 1.969405135414004e-05,
+      "loss": 0.6913,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3616768213887198,
+      "learning_rate": 1.969343427216272e-05,
+      "loss": 0.6915,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1213424353628918,
+      "learning_rate": 1.969281657818888e-05,
+      "loss": 0.579,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3023375402896349,
+      "learning_rate": 1.9692198272257523e-05,
+      "loss": 0.6318,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2886569685708371,
+      "learning_rate": 1.969157935440768e-05,
+      "loss": 0.6205,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2081973668037393,
+      "learning_rate": 1.9690959824678436e-05,
+      "loss": 0.6209,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4683630413264368,
+      "learning_rate": 1.9690339683108894e-05,
+      "loss": 0.6566,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3190817881106844,
+      "learning_rate": 1.9689718929738215e-05,
+      "loss": 0.6452,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4568964100642856,
+      "learning_rate": 1.968909756460559e-05,
+      "loss": 0.618,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3458943224257083,
+      "learning_rate": 1.968847558775024e-05,
+      "loss": 0.6899,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3247938229636835,
+      "learning_rate": 1.9687852999211446e-05,
+      "loss": 0.6129,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2653562001308303,
+      "learning_rate": 1.9687229799028506e-05,
+      "loss": 0.6127,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3502354805046861,
+      "learning_rate": 1.968660598724077e-05,
+      "loss": 0.5831,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3013737334422124,
+      "learning_rate": 1.9685981563887623e-05,
+      "loss": 0.6494,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2750670882881496,
+      "learning_rate": 1.9685356529008487e-05,
+      "loss": 0.6792,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.363217109111308,
+      "learning_rate": 1.9684730882642822e-05,
+      "loss": 0.6512,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4343266497672076,
+      "learning_rate": 1.9684104624830128e-05,
+      "loss": 0.6691,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.427636181002575,
+      "learning_rate": 1.9683477755609946e-05,
+      "loss": 0.6198,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3264707248245353,
+      "learning_rate": 1.9682850275021858e-05,
+      "loss": 0.6036,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.327145114021547,
+      "learning_rate": 1.968222218310547e-05,
+      "loss": 0.6261,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4332285735563004,
+      "learning_rate": 1.9681593479900444e-05,
+      "loss": 0.6474,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3317978282465646,
+      "learning_rate": 1.9680964165446475e-05,
+      "loss": 0.6205,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3282284864610847,
+      "learning_rate": 1.9680334239783286e-05,
+      "loss": 0.6482,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2872462105842508,
+      "learning_rate": 1.9679703702950653e-05,
+      "loss": 0.5826,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.317884074115748,
+      "learning_rate": 1.9679072554988387e-05,
+      "loss": 0.6478,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.477304100823874,
+      "learning_rate": 1.9678440795936332e-05,
+      "loss": 0.727,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2830047686362602,
+      "learning_rate": 1.9677808425834374e-05,
+      "loss": 0.6241,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4408850412410226,
+      "learning_rate": 1.967717544472244e-05,
+      "loss": 0.7432,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.374670162520849,
+      "learning_rate": 1.967654185264049e-05,
+      "loss": 0.7134,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4369771255233479,
+      "learning_rate": 1.9675907649628532e-05,
+      "loss": 0.6893,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.370569938740028,
+      "learning_rate": 1.9675272835726604e-05,
+      "loss": 0.6755,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3160468270194257,
+      "learning_rate": 1.967463741097478e-05,
+      "loss": 0.6197,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2821117852374126,
+      "learning_rate": 1.9674001375413182e-05,
+      "loss": 0.6588,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3724511972029891,
+      "learning_rate": 1.9673364729081968e-05,
+      "loss": 0.6658,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3909350703159726,
+      "learning_rate": 1.967272747202133e-05,
+      "loss": 0.6973,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4262784744081567,
+      "learning_rate": 1.9672089604271503e-05,
+      "loss": 0.6834,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.402619358702424,
+      "learning_rate": 1.9671451125872757e-05,
+      "loss": 0.6341,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2839877489194766,
+      "learning_rate": 1.9670812036865402e-05,
+      "loss": 0.5866,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3311213184900068,
+      "learning_rate": 1.967017233728979e-05,
+      "loss": 0.7219,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.291405092436173,
+      "learning_rate": 1.9669532027186308e-05,
+      "loss": 0.666,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3878211998532612,
+      "learning_rate": 1.966889110659538e-05,
+      "loss": 0.6766,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3578105123173922,
+      "learning_rate": 1.9668249575557468e-05,
+      "loss": 0.6407,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5231094398376817,
+      "learning_rate": 1.966760743411308e-05,
+      "loss": 0.701,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.33094488238159,
+      "learning_rate": 1.9666964682302757e-05,
+      "loss": 0.6604,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4229809159629738,
+      "learning_rate": 1.9666321320167083e-05,
+      "loss": 0.6916,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.353501312928819,
+      "learning_rate": 1.9665677347746667e-05,
+      "loss": 0.6485,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4050817511317932,
+      "learning_rate": 1.9665032765082174e-05,
+      "loss": 0.6948,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3017059338728252,
+      "learning_rate": 1.96643875722143e-05,
+      "loss": 0.6522,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4904558609624012,
+      "learning_rate": 1.966374176918377e-05,
+      "loss": 0.666,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2718702630182106,
+      "learning_rate": 1.966309535603137e-05,
+      "loss": 0.6637,
+      "step": 1423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3190037035857545,
+      "learning_rate": 1.9662448332797905e-05,
+      "loss": 0.6748,
+      "step": 1424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2997543818184787,
+      "learning_rate": 1.9661800699524218e-05,
+      "loss": 0.6677,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.263777243512033,
+      "learning_rate": 1.966115245625121e-05,
+      "loss": 0.6109,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.397406810935863,
+      "learning_rate": 1.96605036030198e-05,
+      "loss": 0.6861,
+      "step": 1427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.34304680278459,
+      "learning_rate": 1.9659854139870963e-05,
+      "loss": 0.6652,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3773783575027967,
+      "learning_rate": 1.965920406684569e-05,
+      "loss": 0.6398,
+      "step": 1429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2313327722756415,
+      "learning_rate": 1.965855338398503e-05,
+      "loss": 0.5613,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2920659996394597,
+      "learning_rate": 1.965790209133006e-05,
+      "loss": 0.5961,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3497886704234359,
+      "learning_rate": 1.9657250188921906e-05,
+      "loss": 0.6261,
+      "step": 1432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2567181772877916,
+      "learning_rate": 1.9656597676801726e-05,
+      "loss": 0.5708,
+      "step": 1433
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.28523026922812,
+      "learning_rate": 1.9655944555010705e-05,
+      "loss": 0.6334,
+      "step": 1434
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3842284717039726,
+      "learning_rate": 1.9655290823590095e-05,
+      "loss": 0.6412,
+      "step": 1435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5027082154882883,
+      "learning_rate": 1.9654636482581157e-05,
+      "loss": 0.6968,
+      "step": 1436
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2456872450401624,
+      "learning_rate": 1.9653981532025206e-05,
+      "loss": 0.6489,
+      "step": 1437
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4398361793687466,
+      "learning_rate": 1.9653325971963594e-05,
+      "loss": 0.6881,
+      "step": 1438
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.363192448825361,
+      "learning_rate": 1.9652669802437704e-05,
+      "loss": 0.6535,
+      "step": 1439
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5079080818201034,
+      "learning_rate": 1.9652013023488972e-05,
+      "loss": 0.7191,
+      "step": 1440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3701218423472024,
+      "learning_rate": 1.9651355635158863e-05,
+      "loss": 0.6231,
+      "step": 1441
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4454980679541667,
+      "learning_rate": 1.965069763748887e-05,
+      "loss": 0.7062,
+      "step": 1442
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4891703995862617,
+      "learning_rate": 1.9650039030520552e-05,
+      "loss": 0.6624,
+      "step": 1443
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3884134172801812,
+      "learning_rate": 1.964937981429548e-05,
+      "loss": 0.6836,
+      "step": 1444
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.370083820048184,
+      "learning_rate": 1.9648719988855278e-05,
+      "loss": 0.6647,
+      "step": 1445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4064895002079651,
+      "learning_rate": 1.9648059554241596e-05,
+      "loss": 0.6579,
+      "step": 1446
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3813042470368648,
+      "learning_rate": 1.9647398510496144e-05,
+      "loss": 0.6635,
+      "step": 1447
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5169813241934325,
+      "learning_rate": 1.9646736857660644e-05,
+      "loss": 0.6605,
+      "step": 1448
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.326361663676536,
+      "learning_rate": 1.964607459577688e-05,
+      "loss": 0.6462,
+      "step": 1449
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4212967252138153,
+      "learning_rate": 1.964541172488666e-05,
+      "loss": 0.6866,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4149328261697203,
+      "learning_rate": 1.964474824503183e-05,
+      "loss": 0.6247,
+      "step": 1451
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3792144171678165,
+      "learning_rate": 1.9644084156254285e-05,
+      "loss": 0.5731,
+      "step": 1452
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.379319774661222,
+      "learning_rate": 1.9643419458595948e-05,
+      "loss": 0.6721,
+      "step": 1453
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2561116061677613,
+      "learning_rate": 1.964275415209879e-05,
+      "loss": 0.6117,
+      "step": 1454
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3790784085356942,
+      "learning_rate": 1.964208823680481e-05,
+      "loss": 0.6504,
+      "step": 1455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.297836544733584,
+      "learning_rate": 1.9641421712756055e-05,
+      "loss": 0.6207,
+      "step": 1456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2937489716898936,
+      "learning_rate": 1.9640754579994607e-05,
+      "loss": 0.6751,
+      "step": 1457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2338673960324187,
+      "learning_rate": 1.9640086838562577e-05,
+      "loss": 0.6105,
+      "step": 1458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3032162074089928,
+      "learning_rate": 1.9639418488502132e-05,
+      "loss": 0.6655,
+      "step": 1459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5097029266244295,
+      "learning_rate": 1.9638749529855463e-05,
+      "loss": 0.672,
+      "step": 1460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.33194649614095,
+      "learning_rate": 1.9638079962664807e-05,
+      "loss": 0.6004,
+      "step": 1461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4529644456991018,
+      "learning_rate": 1.963740978697244e-05,
+      "loss": 0.6921,
+      "step": 1462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3697615887297365,
+      "learning_rate": 1.9636739002820665e-05,
+      "loss": 0.6528,
+      "step": 1463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4559886157513002,
+      "learning_rate": 1.9636067610251842e-05,
+      "loss": 0.7375,
+      "step": 1464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4152314639081294,
+      "learning_rate": 1.9635395609308354e-05,
+      "loss": 0.7021,
+      "step": 1465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3959513894255413,
+      "learning_rate": 1.963472300003263e-05,
+      "loss": 0.6407,
+      "step": 1466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3531217456100126,
+      "learning_rate": 1.9634049782467132e-05,
+      "loss": 0.6784,
+      "step": 1467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3059475233684217,
+      "learning_rate": 1.9633375956654368e-05,
+      "loss": 0.6873,
+      "step": 1468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3286261566671012,
+      "learning_rate": 1.9632701522636878e-05,
+      "loss": 0.6732,
+      "step": 1469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3085873276282465,
+      "learning_rate": 1.9632026480457243e-05,
+      "loss": 0.6367,
+      "step": 1470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2821963002337313,
+      "learning_rate": 1.963135083015808e-05,
+      "loss": 0.6455,
+      "step": 1471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4285298647964901,
+      "learning_rate": 1.963067457178205e-05,
+      "loss": 0.7056,
+      "step": 1472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3905576625336742,
+      "learning_rate": 1.9629997705371845e-05,
+      "loss": 0.5912,
+      "step": 1473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.318258052747074,
+      "learning_rate": 1.96293202309702e-05,
+      "loss": 0.7171,
+      "step": 1474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.41699597327307,
+      "learning_rate": 1.962864214861989e-05,
+      "loss": 0.6417,
+      "step": 1475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2461805163941169,
+      "learning_rate": 1.9627963458363722e-05,
+      "loss": 0.5996,
+      "step": 1476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3295378855213416,
+      "learning_rate": 1.962728416024455e-05,
+      "loss": 0.6438,
+      "step": 1477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3806184507243149,
+      "learning_rate": 1.9626604254305258e-05,
+      "loss": 0.6045,
+      "step": 1478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4109708171641018,
+      "learning_rate": 1.9625923740588767e-05,
+      "loss": 0.6211,
+      "step": 1479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4741973503431585,
+      "learning_rate": 1.9625242619138053e-05,
+      "loss": 0.7101,
+      "step": 1480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5061128356021916,
+      "learning_rate": 1.962456088999611e-05,
+      "loss": 0.642,
+      "step": 1481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3892191520727235,
+      "learning_rate": 1.9623878553205986e-05,
+      "loss": 0.6583,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2544597224204415,
+      "learning_rate": 1.9623195608810753e-05,
+      "loss": 0.6542,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4103987213225255,
+      "learning_rate": 1.9622512056853533e-05,
+      "loss": 0.6287,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2931652092708912,
+      "learning_rate": 1.962182789737748e-05,
+      "loss": 0.6521,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.405806535097205,
+      "learning_rate": 1.962114313042579e-05,
+      "loss": 0.7069,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3555031807783708,
+      "learning_rate": 1.9620457756041695e-05,
+      "loss": 0.6509,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2078097897359783,
+      "learning_rate": 1.961977177426847e-05,
+      "loss": 0.6376,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3442926420117547,
+      "learning_rate": 1.9619085185149417e-05,
+      "loss": 0.6412,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.514891614152846,
+      "learning_rate": 1.9618397988727892e-05,
+      "loss": 0.7223,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2858533736715312,
+      "learning_rate": 1.9617710185047274e-05,
+      "loss": 0.66,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4054810859007343,
+      "learning_rate": 1.9617021774150996e-05,
+      "loss": 0.7262,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4371497515361626,
+      "learning_rate": 1.961633275608251e-05,
+      "loss": 0.7114,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2273913456206889,
+      "learning_rate": 1.9615643130885327e-05,
+      "loss": 0.6181,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.267532089173306,
+      "learning_rate": 1.9614952898602983e-05,
+      "loss": 0.6302,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3370799929271104,
+      "learning_rate": 1.9614262059279054e-05,
+      "loss": 0.6239,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4962116722864225,
+      "learning_rate": 1.9613570612957164e-05,
+      "loss": 0.686,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2704701399733982,
+      "learning_rate": 1.9612878559680953e-05,
+      "loss": 0.6545,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1557092690986264,
+      "learning_rate": 1.9612185899494127e-05,
+      "loss": 0.6234,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4162674883766881,
+      "learning_rate": 1.9611492632440415e-05,
+      "loss": 0.6503,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4207065518253712,
+      "learning_rate": 1.961079875856358e-05,
+      "loss": 0.69,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2116551395446886,
+      "learning_rate": 1.961010427790744e-05,
+      "loss": 0.6274,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3271790670042098,
+      "learning_rate": 1.960940919051583e-05,
+      "loss": 0.6931,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4489016121632021,
+      "learning_rate": 1.9608713496432637e-05,
+      "loss": 0.672,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.43767007568,
+      "learning_rate": 1.9608017195701795e-05,
+      "loss": 0.6606,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3455544709371796,
+      "learning_rate": 1.960732028836725e-05,
+      "loss": 0.6673,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3500249418850947,
+      "learning_rate": 1.960662277447301e-05,
+      "loss": 0.6585,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4347901717905132,
+      "learning_rate": 1.960592465406311e-05,
+      "loss": 0.7402,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.393893953766652,
+      "learning_rate": 1.960522592718163e-05,
+      "loss": 0.6408,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0830572889221242,
+      "learning_rate": 1.9604526593872678e-05,
+      "loss": 0.5447,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2706019654745004,
+      "learning_rate": 1.960382665418041e-05,
+      "loss": 0.6652,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2725235169041942,
+      "learning_rate": 1.9603126108149014e-05,
+      "loss": 0.6438,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2505317987260216,
+      "learning_rate": 1.9602424955822725e-05,
+      "loss": 0.5812,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2459194814863555,
+      "learning_rate": 1.9601723197245803e-05,
+      "loss": 0.6672,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3984968716256014,
+      "learning_rate": 1.960102083246256e-05,
+      "loss": 0.7122,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.450813505916241,
+      "learning_rate": 1.9600317861517335e-05,
+      "loss": 0.6872,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3646439351998105,
+      "learning_rate": 1.959961428445451e-05,
+      "loss": 0.6125,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2954163678793544,
+      "learning_rate": 1.9598910101318514e-05,
+      "loss": 0.5886,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.25640629417667,
+      "learning_rate": 1.9598205312153795e-05,
+      "loss": 0.6207,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4745581207462595,
+      "learning_rate": 1.9597499917004856e-05,
+      "loss": 0.7358,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.319866172696702,
+      "learning_rate": 1.9596793915916227e-05,
+      "loss": 0.6598,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3102367051135526,
+      "learning_rate": 1.959608730893249e-05,
+      "loss": 0.6806,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2986111796759294,
+      "learning_rate": 1.959538009609825e-05,
+      "loss": 0.6256,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.312147865287336,
+      "learning_rate": 1.959467227745816e-05,
+      "loss": 0.6063,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2882306738480376,
+      "learning_rate": 1.9593963853056905e-05,
+      "loss": 0.6839,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4693887011734195,
+      "learning_rate": 1.9593254822939215e-05,
+      "loss": 0.6933,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.400339846574915,
+      "learning_rate": 1.9592545187149855e-05,
+      "loss": 0.6879,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3867996836940029,
+      "learning_rate": 1.9591834945733625e-05,
+      "loss": 0.6391,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4110075264722555,
+      "learning_rate": 1.9591124098735363e-05,
+      "loss": 0.6457,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3716469977132435,
+      "learning_rate": 1.959041264619996e-05,
+      "loss": 0.6833,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3963892337427741,
+      "learning_rate": 1.958970058817232e-05,
+      "loss": 0.639,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5121028126421716,
+      "learning_rate": 1.958898792469741e-05,
+      "loss": 0.7052,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5256500609643444,
+      "learning_rate": 1.958827465582022e-05,
+      "loss": 0.7058,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4835974815285566,
+      "learning_rate": 1.958756078158578e-05,
+      "loss": 0.7067,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.349144111517021,
+      "learning_rate": 1.958684630203916e-05,
+      "loss": 0.624,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4271914680146813,
+      "learning_rate": 1.958613121722548e-05,
+      "loss": 0.663,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3336491757067224,
+      "learning_rate": 1.958541552718987e-05,
+      "loss": 0.6791,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5672743525071198,
+      "learning_rate": 1.9584699231977526e-05,
+      "loss": 0.633,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2150944392859622,
+      "learning_rate": 1.9583982331633668e-05,
+      "loss": 0.5963,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2103470377939822,
+      "learning_rate": 1.9583264826203564e-05,
+      "loss": 0.5991,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.49303671753505,
+      "learning_rate": 1.9582546715732503e-05,
+      "loss": 0.6851,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1446403477460383,
+      "learning_rate": 1.958182800026583e-05,
+      "loss": 0.576,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3180844620774201,
+      "learning_rate": 1.9581108679848918e-05,
+      "loss": 0.623,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2114147107357678,
+      "learning_rate": 1.9580388754527185e-05,
+      "loss": 0.598,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1869258998665362,
+      "learning_rate": 1.957966822434608e-05,
+      "loss": 0.6084,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2582476319355758,
+      "learning_rate": 1.9578947089351095e-05,
+      "loss": 0.6042,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2811762160478832,
+      "learning_rate": 1.9578225349587763e-05,
+      "loss": 0.5874,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3432276398008005,
+      "learning_rate": 1.9577503005101644e-05,
+      "loss": 0.6726,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3882739735049308,
+      "learning_rate": 1.9576780055938348e-05,
+      "loss": 0.6434,
+      "step": 1549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3250526345792901,
+      "learning_rate": 1.957605650214352e-05,
+      "loss": 0.692,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3034620183130825,
+      "learning_rate": 1.9575332343762832e-05,
+      "loss": 0.6672,
+      "step": 1551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2811228060972017,
+      "learning_rate": 1.9574607580842018e-05,
+      "loss": 0.6272,
+      "step": 1552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3208527672879262,
+      "learning_rate": 1.9573882213426824e-05,
+      "loss": 0.5954,
+      "step": 1553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2306002592018632,
+      "learning_rate": 1.9573156241563053e-05,
+      "loss": 0.6036,
+      "step": 1554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2603278273282519,
+      "learning_rate": 1.957242966529654e-05,
+      "loss": 0.6039,
+      "step": 1555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2573860819884088,
+      "learning_rate": 1.9571702484673153e-05,
+      "loss": 0.6227,
+      "step": 1556
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2413642602520683,
+      "learning_rate": 1.9570974699738805e-05,
+      "loss": 0.5891,
+      "step": 1557
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3526493179057293,
+      "learning_rate": 1.9570246310539443e-05,
+      "loss": 0.6823,
+      "step": 1558
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4106626578777612,
+      "learning_rate": 1.9569517317121058e-05,
+      "loss": 0.6279,
+      "step": 1559
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.233501656533108,
+      "learning_rate": 1.9568787719529673e-05,
+      "loss": 0.6215,
+      "step": 1560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4274298347024477,
+      "learning_rate": 1.9568057517811345e-05,
+      "loss": 0.674,
+      "step": 1561
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2834856485828359,
+      "learning_rate": 1.9567326712012188e-05,
+      "loss": 0.6489,
+      "step": 1562
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3149857370161504,
+      "learning_rate": 1.956659530217833e-05,
+      "loss": 0.6282,
+      "step": 1563
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3475865000490472,
+      "learning_rate": 1.9565863288355955e-05,
+      "loss": 0.6411,
+      "step": 1564
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2486982242821192,
+      "learning_rate": 1.956513067059128e-05,
+      "loss": 0.647,
+      "step": 1565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.347892629928924,
+      "learning_rate": 1.9564397448930552e-05,
+      "loss": 0.7104,
+      "step": 1566
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2879110568979681,
+      "learning_rate": 1.9563663623420067e-05,
+      "loss": 0.6506,
+      "step": 1567
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.301153914305594,
+      "learning_rate": 1.956292919410616e-05,
+      "loss": 0.6711,
+      "step": 1568
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3308354712898856,
+      "learning_rate": 1.9562194161035194e-05,
+      "loss": 0.6517,
+      "step": 1569
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3553894635408312,
+      "learning_rate": 1.956145852425357e-05,
+      "loss": 0.616,
+      "step": 1570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3778268193075993,
+      "learning_rate": 1.9560722283807744e-05,
+      "loss": 0.6846,
+      "step": 1571
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.38099633040828,
+      "learning_rate": 1.955998543974419e-05,
+      "loss": 0.6395,
+      "step": 1572
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.275608762631205,
+      "learning_rate": 1.955924799210944e-05,
+      "loss": 0.5865,
+      "step": 1573
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2699009283726723,
+      "learning_rate": 1.9558509940950036e-05,
+      "loss": 0.6153,
+      "step": 1574
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2834906640569594,
+      "learning_rate": 1.955777128631259e-05,
+      "loss": 0.6268,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1848884777916726,
+      "learning_rate": 1.9557032028243728e-05,
+      "loss": 0.6072,
+      "step": 1576
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4470417297399003,
+      "learning_rate": 1.9556292166790128e-05,
+      "loss": 0.6804,
+      "step": 1577
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3033755896239438,
+      "learning_rate": 1.9555551701998496e-05,
+      "loss": 0.5977,
+      "step": 1578
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2017973473151253,
+      "learning_rate": 1.9554810633915587e-05,
+      "loss": 0.5935,
+      "step": 1579
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2440387199024459,
+      "learning_rate": 1.955406896258819e-05,
+      "loss": 0.6438,
+      "step": 1580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.308410905048871,
+      "learning_rate": 1.955332668806312e-05,
+      "loss": 0.649,
+      "step": 1581
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4653000591365797,
+      "learning_rate": 1.9552583810387253e-05,
+      "loss": 0.6331,
+      "step": 1582
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.303129076322405,
+      "learning_rate": 1.9551840329607483e-05,
+      "loss": 0.6694,
+      "step": 1583
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.287249822290956,
+      "learning_rate": 1.9551096245770752e-05,
+      "loss": 0.6592,
+      "step": 1584
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.329040167159197,
+      "learning_rate": 1.9550351558924036e-05,
+      "loss": 0.7047,
+      "step": 1585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3599395894910644,
+      "learning_rate": 1.9549606269114358e-05,
+      "loss": 0.5875,
+      "step": 1586
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.360770112527667,
+      "learning_rate": 1.9548860376388762e-05,
+      "loss": 0.6684,
+      "step": 1587
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.401916863968186,
+      "learning_rate": 1.9548113880794348e-05,
+      "loss": 0.7019,
+      "step": 1588
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.432876116256592,
+      "learning_rate": 1.954736678237824e-05,
+      "loss": 0.6843,
+      "step": 1589
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3406597355252898,
+      "learning_rate": 1.954661908118761e-05,
+      "loss": 0.6047,
+      "step": 1590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3159802027549536,
+      "learning_rate": 1.9545870777269664e-05,
+      "loss": 0.6297,
+      "step": 1591
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3958866574174782,
+      "learning_rate": 1.9545121870671642e-05,
+      "loss": 0.6356,
+      "step": 1592
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3212703413105757,
+      "learning_rate": 1.9544372361440836e-05,
+      "loss": 0.676,
+      "step": 1593
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4324014055231864,
+      "learning_rate": 1.9543622249624557e-05,
+      "loss": 0.7143,
+      "step": 1594
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3510361898051917,
+      "learning_rate": 1.9542871535270168e-05,
+      "loss": 0.5936,
+      "step": 1595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.433319481331774,
+      "learning_rate": 1.9542120218425062e-05,
+      "loss": 0.7274,
+      "step": 1596
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4207225782495838,
+      "learning_rate": 1.954136829913668e-05,
+      "loss": 0.7697,
+      "step": 1597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.172925452859515,
+      "learning_rate": 1.954061577745249e-05,
+      "loss": 0.6404,
+      "step": 1598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3739292570695558,
+      "learning_rate": 1.9539862653419998e-05,
+      "loss": 0.707,
+      "step": 1599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2791762433433942,
+      "learning_rate": 1.9539108927086762e-05,
+      "loss": 0.6349,
+      "step": 1600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.41234982291409,
+      "learning_rate": 1.9538354598500358e-05,
+      "loss": 0.6279,
+      "step": 1601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3472062396164377,
+      "learning_rate": 1.9537599667708423e-05,
+      "loss": 0.6537,
+      "step": 1602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3598745020295593,
+      "learning_rate": 1.953684413475861e-05,
+      "loss": 0.6575,
+      "step": 1603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3492193030978572,
+      "learning_rate": 1.9536087999698624e-05,
+      "loss": 0.6921,
+      "step": 1604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3350143366126004,
+      "learning_rate": 1.9535331262576203e-05,
+      "loss": 0.6351,
+      "step": 1605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3007999453189176,
+      "learning_rate": 1.9534573923439124e-05,
+      "loss": 0.6497,
+      "step": 1606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4242711261068313,
+      "learning_rate": 1.95338159823352e-05,
+      "loss": 0.6587,
+      "step": 1607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3295646942483512,
+      "learning_rate": 1.9533057439312286e-05,
+      "loss": 0.6443,
+      "step": 1608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.285179253925957,
+      "learning_rate": 1.9532298294418272e-05,
+      "loss": 0.66,
+      "step": 1609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4129158133786655,
+      "learning_rate": 1.9531538547701087e-05,
+      "loss": 0.6448,
+      "step": 1610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3419376393850617,
+      "learning_rate": 1.9530778199208698e-05,
+      "loss": 0.6247,
+      "step": 1611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3834339189536926,
+      "learning_rate": 1.9530017248989107e-05,
+      "loss": 0.6187,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4451985082264414,
+      "learning_rate": 1.9529255697090358e-05,
+      "loss": 0.6755,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.349543347655055,
+      "learning_rate": 1.9528493543560533e-05,
+      "loss": 0.6281,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3757705696726843,
+      "learning_rate": 1.9527730788447752e-05,
+      "loss": 0.6459,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2725048277188138,
+      "learning_rate": 1.952696743180017e-05,
+      "loss": 0.6539,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4875827862438296,
+      "learning_rate": 1.952620347366598e-05,
+      "loss": 0.6947,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.494804284799022,
+      "learning_rate": 1.952543891409341e-05,
+      "loss": 0.6922,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3048225835436218,
+      "learning_rate": 1.9524673753130745e-05,
+      "loss": 0.6317,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.456086780749599,
+      "learning_rate": 1.952390799082628e-05,
+      "loss": 0.717,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3076789505198712,
+      "learning_rate": 1.952314162722837e-05,
+      "loss": 0.67,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2647147025049956,
+      "learning_rate": 1.9522374662385396e-05,
+      "loss": 0.5876,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.495839149584217,
+      "learning_rate": 1.952160709634578e-05,
+      "loss": 0.7636,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2971599219911996,
+      "learning_rate": 1.952083892915798e-05,
+      "loss": 0.6187,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3695968324982255,
+      "learning_rate": 1.95200701608705e-05,
+      "loss": 0.6537,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3637421274351307,
+      "learning_rate": 1.9519300791531874e-05,
+      "loss": 0.6864,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2084629493560946,
+      "learning_rate": 1.9518530821190675e-05,
+      "loss": 0.6336,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3328254099821017,
+      "learning_rate": 1.9517760249895518e-05,
+      "loss": 0.6069,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.226836107194319,
+      "learning_rate": 1.951698907769505e-05,
+      "loss": 0.6174,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2896108674708435,
+      "learning_rate": 1.951621730463796e-05,
+      "loss": 0.6248,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2564568648184522,
+      "learning_rate": 1.9515444930772972e-05,
+      "loss": 0.6375,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3202292963480275,
+      "learning_rate": 1.9514671956148855e-05,
+      "loss": 0.6724,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2474507085507114,
+      "learning_rate": 1.951389838081441e-05,
+      "loss": 0.6047,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5826478109888913,
+      "learning_rate": 1.951312420481847e-05,
+      "loss": 0.7121,
+      "step": 1634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3152930649662653,
+      "learning_rate": 1.951234942820992e-05,
+      "loss": 0.6369,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3218693780441315,
+      "learning_rate": 1.9511574051037672e-05,
+      "loss": 0.6749,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2525439125632163,
+      "learning_rate": 1.9510798073350686e-05,
+      "loss": 0.6181,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.546422487597453,
+      "learning_rate": 1.9510021495197943e-05,
+      "loss": 0.6291,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3831334145794631,
+      "learning_rate": 1.950924431662848e-05,
+      "loss": 0.7096,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2982931082047735,
+      "learning_rate": 1.9508466537691363e-05,
+      "loss": 0.6685,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2439292836006322,
+      "learning_rate": 1.9507688158435693e-05,
+      "loss": 0.6459,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.218063356690392,
+      "learning_rate": 1.950690917891062e-05,
+      "loss": 0.6122,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2613990307746286,
+      "learning_rate": 1.950612959916532e-05,
+      "loss": 0.6282,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2630363646925478,
+      "learning_rate": 1.9505349419249015e-05,
+      "loss": 0.6726,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4247964880996937,
+      "learning_rate": 1.9504568639210956e-05,
+      "loss": 0.6326,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3253855558164094,
+      "learning_rate": 1.9503787259100445e-05,
+      "loss": 0.6852,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4069105504279644,
+      "learning_rate": 1.9503005278966808e-05,
+      "loss": 0.6913,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3431000579321597,
+      "learning_rate": 1.9502222698859422e-05,
+      "loss": 0.6527,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.298649320986588,
+      "learning_rate": 1.950143951882769e-05,
+      "loss": 0.6676,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2262062964783422,
+      "learning_rate": 1.9500655738921058e-05,
+      "loss": 0.5631,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2051690807670927,
+      "learning_rate": 1.9499871359189017e-05,
+      "loss": 0.643,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2912285379581527,
+      "learning_rate": 1.949908637968108e-05,
+      "loss": 0.6451,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.413028106876958,
+      "learning_rate": 1.9498300800446815e-05,
+      "loss": 0.6963,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3135889622126984,
+      "learning_rate": 1.9497514621535815e-05,
+      "loss": 0.6241,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2908234561438021,
+      "learning_rate": 1.9496727842997713e-05,
+      "loss": 0.6443,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2428083487737023,
+      "learning_rate": 1.949594046488219e-05,
+      "loss": 0.6417,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1918627255332315,
+      "learning_rate": 1.949515248723895e-05,
+      "loss": 0.5729,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3364460295152696,
+      "learning_rate": 1.9494363910117745e-05,
+      "loss": 0.6751,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3817888339476168,
+      "learning_rate": 1.9493574733568362e-05,
+      "loss": 0.6299,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.336331850262456,
+      "learning_rate": 1.9492784957640624e-05,
+      "loss": 0.6401,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5394468650506685,
+      "learning_rate": 1.94919945823844e-05,
+      "loss": 0.6854,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3283612658213446,
+      "learning_rate": 1.949120360784958e-05,
+      "loss": 0.6765,
+      "step": 1662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3156635668395658,
+      "learning_rate": 1.9490412034086112e-05,
+      "loss": 0.6036,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.352655310747795,
+      "learning_rate": 1.9489619861143966e-05,
+      "loss": 0.6887,
+      "step": 1664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3882399690809386,
+      "learning_rate": 1.948882708907316e-05,
+      "loss": 0.6455,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4230400594191541,
+      "learning_rate": 1.9488033717923744e-05,
+      "loss": 0.6678,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4792071681678847,
+      "learning_rate": 1.9487239747745807e-05,
+      "loss": 0.6669,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0823281772440976,
+      "learning_rate": 1.9486445178589477e-05,
+      "loss": 0.5852,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.30532944470769,
+      "learning_rate": 1.9485650010504918e-05,
+      "loss": 0.6312,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.207124132452917,
+      "learning_rate": 1.9484854243542336e-05,
+      "loss": 0.5761,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4307943759977249,
+      "learning_rate": 1.948405787775197e-05,
+      "loss": 0.6544,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2773409035560555,
+      "learning_rate": 1.94832609131841e-05,
+      "loss": 0.6917,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3656208326332877,
+      "learning_rate": 1.948246334988904e-05,
+      "loss": 0.6261,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3661389498937764,
+      "learning_rate": 1.9481665187917147e-05,
+      "loss": 0.6416,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3336420248240874,
+      "learning_rate": 1.9480866427318807e-05,
+      "loss": 0.638,
+      "step": 1675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3478522116782916,
+      "learning_rate": 1.9480067068144458e-05,
+      "loss": 0.6766,
+      "step": 1676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3819711140541093,
+      "learning_rate": 1.9479267110444566e-05,
+      "loss": 0.6582,
+      "step": 1677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3340384834006052,
+      "learning_rate": 1.9478466554269633e-05,
+      "loss": 0.5807,
+      "step": 1678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2321214506212914,
+      "learning_rate": 1.94776653996702e-05,
+      "loss": 0.6527,
+      "step": 1679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.42304382910286,
+      "learning_rate": 1.947686364669686e-05,
+      "loss": 0.7393,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.351307883562319,
+      "learning_rate": 1.9476061295400218e-05,
+      "loss": 0.6712,
+      "step": 1681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2371071625458785,
+      "learning_rate": 1.947525834583094e-05,
+      "loss": 0.6414,
+      "step": 1682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3435047724319267,
+      "learning_rate": 1.9474454798039713e-05,
+      "loss": 0.6386,
+      "step": 1683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4039604622335615,
+      "learning_rate": 1.9473650652077275e-05,
+      "loss": 0.6874,
+      "step": 1684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3784077205135106,
+      "learning_rate": 1.9472845907994397e-05,
+      "loss": 0.6332,
+      "step": 1685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2000282860442137,
+      "learning_rate": 1.9472040565841877e-05,
+      "loss": 0.5581,
+      "step": 1686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4381181797494187,
+      "learning_rate": 1.947123462567057e-05,
+      "loss": 0.6511,
+      "step": 1687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.412294747622676,
+      "learning_rate": 1.9470428087531354e-05,
+      "loss": 0.6266,
+      "step": 1688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3290537559998392,
+      "learning_rate": 1.9469620951475154e-05,
+      "loss": 0.6643,
+      "step": 1689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2301033567407134,
+      "learning_rate": 1.9468813217552926e-05,
+      "loss": 0.6454,
+      "step": 1690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2913551518923005,
+      "learning_rate": 1.9468004885815667e-05,
+      "loss": 0.6124,
+      "step": 1691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2024786563018028,
+      "learning_rate": 1.946719595631441e-05,
+      "loss": 0.5772,
+      "step": 1692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4841037703217972,
+      "learning_rate": 1.946638642910023e-05,
+      "loss": 0.6568,
+      "step": 1693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3363512971057134,
+      "learning_rate": 1.9465576304224233e-05,
+      "loss": 0.6735,
+      "step": 1694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3300614545004312,
+      "learning_rate": 1.9464765581737567e-05,
+      "loss": 0.6477,
+      "step": 1695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3342649910313202,
+      "learning_rate": 1.946395426169142e-05,
+      "loss": 0.6188,
+      "step": 1696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.356269709711965,
+      "learning_rate": 1.9463142344137013e-05,
+      "loss": 0.6254,
+      "step": 1697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2737688908846687,
+      "learning_rate": 1.9462329829125604e-05,
+      "loss": 0.6352,
+      "step": 1698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2928672819861524,
+      "learning_rate": 1.9461516716708496e-05,
+      "loss": 0.6746,
+      "step": 1699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2462527852232945,
+      "learning_rate": 1.9460703006937023e-05,
+      "loss": 0.6072,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3600213717436305,
+      "learning_rate": 1.9459888699862555e-05,
+      "loss": 0.6699,
+      "step": 1701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.304672241121708,
+      "learning_rate": 1.9459073795536513e-05,
+      "loss": 0.6614,
+      "step": 1702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.260479344892915,
+      "learning_rate": 1.9458258294010335e-05,
+      "loss": 0.5953,
+      "step": 1703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4027928135425936,
+      "learning_rate": 1.9457442195335514e-05,
+      "loss": 0.6853,
+      "step": 1704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.243047116900357,
+      "learning_rate": 1.945662549956357e-05,
+      "loss": 0.6072,
+      "step": 1705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2468337011548778,
+      "learning_rate": 1.945580820674607e-05,
+      "loss": 0.6176,
+      "step": 1706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4600280542160484,
+      "learning_rate": 1.9454990316934618e-05,
+      "loss": 0.6828,
+      "step": 1707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4361574288889705,
+      "learning_rate": 1.9454171830180842e-05,
+      "loss": 0.6607,
+      "step": 1708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2911049123754939,
+      "learning_rate": 1.945335274653642e-05,
+      "loss": 0.656,
+      "step": 1709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.220834953819128,
+      "learning_rate": 1.9452533066053067e-05,
+      "loss": 0.5929,
+      "step": 1710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2237992474131756,
+      "learning_rate": 1.9451712788782534e-05,
+      "loss": 0.5865,
+      "step": 1711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.386744969079819,
+      "learning_rate": 1.9450891914776605e-05,
+      "loss": 0.6868,
+      "step": 1712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2609752436664927,
+      "learning_rate": 1.9450070444087113e-05,
+      "loss": 0.6056,
+      "step": 1713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2892210429403175,
+      "learning_rate": 1.9449248376765918e-05,
+      "loss": 0.5985,
+      "step": 1714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4446119930628774,
+      "learning_rate": 1.9448425712864917e-05,
+      "loss": 0.6983,
+      "step": 1715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4259015594306956,
+      "learning_rate": 1.9447602452436058e-05,
+      "loss": 0.6322,
+      "step": 1716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.226230503514461,
+      "learning_rate": 1.944677859553131e-05,
+      "loss": 0.6451,
+      "step": 1717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.283456066174709,
+      "learning_rate": 1.944595414220269e-05,
+      "loss": 0.6641,
+      "step": 1718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3642209811962993,
+      "learning_rate": 1.9445129092502253e-05,
+      "loss": 0.6395,
+      "step": 1719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.245708441879969,
+      "learning_rate": 1.9444303446482086e-05,
+      "loss": 0.6451,
+      "step": 1720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3245842353327342,
+      "learning_rate": 1.9443477204194315e-05,
+      "loss": 0.6655,
+      "step": 1721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2715552056854047,
+      "learning_rate": 1.9442650365691102e-05,
+      "loss": 0.6386,
+      "step": 1722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3462189459517546,
+      "learning_rate": 1.944182293102466e-05,
+      "loss": 0.6391,
+      "step": 1723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2734904366004214,
+      "learning_rate": 1.944099490024722e-05,
+      "loss": 0.6149,
+      "step": 1724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3659556043232803,
+      "learning_rate": 1.9440166273411062e-05,
+      "loss": 0.6347,
+      "step": 1725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2292438525949423,
+      "learning_rate": 1.9439337050568504e-05,
+      "loss": 0.5649,
+      "step": 1726
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2587119257149153,
+      "learning_rate": 1.9438507231771897e-05,
+      "loss": 0.6533,
+      "step": 1727
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3552559457439861,
+      "learning_rate": 1.9437676817073635e-05,
+      "loss": 0.6966,
+      "step": 1728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.317781448848758,
+      "learning_rate": 1.9436845806526137e-05,
+      "loss": 0.6053,
+      "step": 1729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.218933091715898,
+      "learning_rate": 1.943601420018188e-05,
+      "loss": 0.6241,
+      "step": 1730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2819330674164047,
+      "learning_rate": 1.9435181998093363e-05,
+      "loss": 0.5859,
+      "step": 1731
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2979205868796004,
+      "learning_rate": 1.9434349200313126e-05,
+      "loss": 0.6189,
+      "step": 1732
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2631881236027205,
+      "learning_rate": 1.9433515806893754e-05,
+      "loss": 0.6205,
+      "step": 1733
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.211010789191571,
+      "learning_rate": 1.9432681817887854e-05,
+      "loss": 0.6586,
+      "step": 1734
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3550807166603973,
+      "learning_rate": 1.943184723334808e-05,
+      "loss": 0.6747,
+      "step": 1735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.325161803460716,
+      "learning_rate": 1.9431012053327135e-05,
+      "loss": 0.6424,
+      "step": 1736
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2578420694796821,
+      "learning_rate": 1.943017627787774e-05,
+      "loss": 0.6117,
+      "step": 1737
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2846426821764276,
+      "learning_rate": 1.9429339907052666e-05,
+      "loss": 0.6048,
+      "step": 1738
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3351920652636022,
+      "learning_rate": 1.942850294090471e-05,
+      "loss": 0.6545,
+      "step": 1739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2275681473253612,
+      "learning_rate": 1.942766537948672e-05,
+      "loss": 0.5942,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.448747913138571,
+      "learning_rate": 1.9426827222851576e-05,
+      "loss": 0.6672,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3551998256423143,
+      "learning_rate": 1.942598847105219e-05,
+      "loss": 0.7107,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2442038144244327,
+      "learning_rate": 1.9425149124141524e-05,
+      "loss": 0.584,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4410549655837623,
+      "learning_rate": 1.9424309182172563e-05,
+      "loss": 0.6376,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3702917116995883,
+      "learning_rate": 1.9423468645198342e-05,
+      "loss": 0.6489,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3973419757274244,
+      "learning_rate": 1.9422627513271927e-05,
+      "loss": 0.6646,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2005454313967445,
+      "learning_rate": 1.9421785786446422e-05,
+      "loss": 0.6064,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4382653065076993,
+      "learning_rate": 1.942094346477497e-05,
+      "loss": 0.633,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2985161657638071,
+      "learning_rate": 1.9420100548310753e-05,
+      "loss": 0.5816,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2276991418873875,
+      "learning_rate": 1.9419257037106983e-05,
+      "loss": 0.6184,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.321233168813278,
+      "learning_rate": 1.9418412931216926e-05,
+      "loss": 0.6215,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2310511104066357,
+      "learning_rate": 1.9417568230693863e-05,
+      "loss": 0.6148,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.201918951174184,
+      "learning_rate": 1.941672293559113e-05,
+      "loss": 0.6409,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2820192679357278,
+      "learning_rate": 1.9415877045962093e-05,
+      "loss": 0.6141,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4418860983287094,
+      "learning_rate": 1.941503056186016e-05,
+      "loss": 0.6538,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5504665103409105,
+      "learning_rate": 1.9414183483338777e-05,
+      "loss": 0.7011,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3230350121055532,
+      "learning_rate": 1.9413335810451412e-05,
+      "loss": 0.6637,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2218971211413936,
+      "learning_rate": 1.9412487543251596e-05,
+      "loss": 0.701,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2913811378322062,
+      "learning_rate": 1.941163868179288e-05,
+      "loss": 0.6714,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2373744399981528,
+      "learning_rate": 1.9410789226128853e-05,
+      "loss": 0.6462,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.385578385095075,
+      "learning_rate": 1.940993917631315e-05,
+      "loss": 0.6733,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.385045420111922,
+      "learning_rate": 1.9409088532399436e-05,
+      "loss": 0.6381,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2244025486197705,
+      "learning_rate": 1.9408237294441422e-05,
+      "loss": 0.6123,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3612472969918086,
+      "learning_rate": 1.9407385462492846e-05,
+      "loss": 0.6183,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3186129403867841,
+      "learning_rate": 1.9406533036607488e-05,
+      "loss": 0.6606,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2598968674065567,
+      "learning_rate": 1.940568001683917e-05,
+      "loss": 0.6348,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3952888119385922,
+      "learning_rate": 1.9404826403241746e-05,
+      "loss": 0.6662,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.206270589306881,
+      "learning_rate": 1.940397219586911e-05,
+      "loss": 0.5745,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2370139294474551,
+      "learning_rate": 1.940311739477519e-05,
+      "loss": 0.591,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3182917824831164,
+      "learning_rate": 1.940226200001395e-05,
+      "loss": 0.6754,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.412846585178781,
+      "learning_rate": 1.940140601163941e-05,
+      "loss": 0.6588,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3167438470375483,
+      "learning_rate": 1.9400549429705597e-05,
+      "loss": 0.6503,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3200466633264807,
+      "learning_rate": 1.93996922542666e-05,
+      "loss": 0.6787,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.230184564621934,
+      "learning_rate": 1.9398834485376534e-05,
+      "loss": 0.6585,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.303266882547969,
+      "learning_rate": 1.9397976123089558e-05,
+      "loss": 0.6916,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.202594541057109,
+      "learning_rate": 1.9397117167459858e-05,
+      "loss": 0.6339,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2912817145530682,
+      "learning_rate": 1.939625761854167e-05,
+      "loss": 0.6583,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2899321425872985,
+      "learning_rate": 1.9395397476389265e-05,
+      "loss": 0.5833,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1868743754260211,
+      "learning_rate": 1.939453674105694e-05,
+      "loss": 0.6209,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3231568256167092,
+      "learning_rate": 1.9393675412599037e-05,
+      "loss": 0.6849,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2318035333767627,
+      "learning_rate": 1.9392813491069944e-05,
+      "loss": 0.5975,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.35401319465519,
+      "learning_rate": 1.9391950976524075e-05,
+      "loss": 0.6275,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3256753657807507,
+      "learning_rate": 1.9391087869015884e-05,
+      "loss": 0.6601,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.431278365074837,
+      "learning_rate": 1.9390224168599864e-05,
+      "loss": 0.7256,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3296045028657328,
+      "learning_rate": 1.9389359875330548e-05,
+      "loss": 0.6998,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3195173589179643,
+      "learning_rate": 1.9388494989262498e-05,
+      "loss": 0.6338,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2677220055692577,
+      "learning_rate": 1.9387629510450318e-05,
+      "loss": 0.6223,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2939126525211937,
+      "learning_rate": 1.9386763438948656e-05,
+      "loss": 0.6535,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.379490153281343,
+      "learning_rate": 1.938589677481219e-05,
+      "loss": 0.6291,
+      "step": 1789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3741811568261744,
+      "learning_rate": 1.9385029518095635e-05,
+      "loss": 0.6691,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2312748282006032,
+      "learning_rate": 1.9384161668853746e-05,
+      "loss": 0.5783,
+      "step": 1791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3642264426030555,
+      "learning_rate": 1.9383293227141316e-05,
+      "loss": 0.6339,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1622154892630587,
+      "learning_rate": 1.9382424193013172e-05,
+      "loss": 0.5877,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3390573134260653,
+      "learning_rate": 1.938155456652418e-05,
+      "loss": 0.627,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3374954419637717,
+      "learning_rate": 1.938068434772925e-05,
+      "loss": 0.6592,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3315944458845368,
+      "learning_rate": 1.9379813536683314e-05,
+      "loss": 0.709,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3022277395916675,
+      "learning_rate": 1.9378942133441357e-05,
+      "loss": 0.683,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.357890008573226,
+      "learning_rate": 1.9378070138058396e-05,
+      "loss": 0.6939,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3082880830330057,
+      "learning_rate": 1.9377197550589484e-05,
+      "loss": 0.6154,
+      "step": 1799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3562562617144207,
+      "learning_rate": 1.9376324371089707e-05,
+      "loss": 0.6646,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3542960863135862,
+      "learning_rate": 1.93754505996142e-05,
+      "loss": 0.672,
+      "step": 1801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2775218501154013,
+      "learning_rate": 1.9374576236218123e-05,
+      "loss": 0.631,
+      "step": 1802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1758775639928287,
+      "learning_rate": 1.9373701280956685e-05,
+      "loss": 0.5294,
+      "step": 1803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2517750058397301,
+      "learning_rate": 1.937282573388512e-05,
+      "loss": 0.6509,
+      "step": 1804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1802088109840623,
+      "learning_rate": 1.937194959505871e-05,
+      "loss": 0.6129,
+      "step": 1805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3102995273813116,
+      "learning_rate": 1.937107286453277e-05,
+      "loss": 0.6104,
+      "step": 1806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3008657891922533,
+      "learning_rate": 1.9370195542362653e-05,
+      "loss": 0.6734,
+      "step": 1807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2025573183860743,
+      "learning_rate": 1.9369317628603744e-05,
+      "loss": 0.599,
+      "step": 1808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3371318363907538,
+      "learning_rate": 1.936843912331148e-05,
+      "loss": 0.6318,
+      "step": 1809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.284488678516641,
+      "learning_rate": 1.936756002654131e-05,
+      "loss": 0.6081,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.217664577632365,
+      "learning_rate": 1.9366680338348755e-05,
+      "loss": 0.5891,
+      "step": 1811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.186817876757914,
+      "learning_rate": 1.9365800058789338e-05,
+      "loss": 0.5698,
+      "step": 1812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2439156752613698,
+      "learning_rate": 1.9364919187918644e-05,
+      "loss": 0.6373,
+      "step": 1813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3006422638744117,
+      "learning_rate": 1.9364037725792287e-05,
+      "loss": 0.6787,
+      "step": 1814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2967798588795363,
+      "learning_rate": 1.9363155672465913e-05,
+      "loss": 0.6798,
+      "step": 1815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2913093636706705,
+      "learning_rate": 1.9362273027995217e-05,
+      "loss": 0.585,
+      "step": 1816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2658277749464204,
+      "learning_rate": 1.9361389792435918e-05,
+      "loss": 0.6483,
+      "step": 1817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2531786081924103,
+      "learning_rate": 1.9360505965843784e-05,
+      "loss": 0.6083,
+      "step": 1818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1847368517965333,
+      "learning_rate": 1.9359621548274616e-05,
+      "loss": 0.6145,
+      "step": 1819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2516421022472517,
+      "learning_rate": 1.935873653978425e-05,
+      "loss": 0.6707,
+      "step": 1820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3681852786960234,
+      "learning_rate": 1.935785094042856e-05,
+      "loss": 0.697,
+      "step": 1821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.351966136130191,
+      "learning_rate": 1.9356964750263464e-05,
+      "loss": 0.633,
+      "step": 1822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2821431187296615,
+      "learning_rate": 1.9356077969344904e-05,
+      "loss": 0.6545,
+      "step": 1823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.31030953499522,
+      "learning_rate": 1.9355190597728874e-05,
+      "loss": 0.6607,
+      "step": 1824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2535364669346227,
+      "learning_rate": 1.9354302635471392e-05,
+      "loss": 0.5522,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3087551643685043,
+      "learning_rate": 1.9353414082628523e-05,
+      "loss": 0.6121,
+      "step": 1826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4113829288775175,
+      "learning_rate": 1.9352524939256367e-05,
+      "loss": 0.637,
+      "step": 1827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4263114883849193,
+      "learning_rate": 1.935163520541106e-05,
+      "loss": 0.63,
+      "step": 1828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2431996858082133,
+      "learning_rate": 1.935074488114877e-05,
+      "loss": 0.6011,
+      "step": 1829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3271925401692968,
+      "learning_rate": 1.9349853966525715e-05,
+      "loss": 0.7236,
+      "step": 1830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2365950406265822,
+      "learning_rate": 1.934896246159814e-05,
+      "loss": 0.64,
+      "step": 1831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2583187814001624,
+      "learning_rate": 1.9348070366422333e-05,
+      "loss": 0.6623,
+      "step": 1832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2761617902410398,
+      "learning_rate": 1.934717768105461e-05,
+      "loss": 0.5737,
+      "step": 1833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3179462152803372,
+      "learning_rate": 1.9346284405551333e-05,
+      "loss": 0.6379,
+      "step": 1834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2759810712271338,
+      "learning_rate": 1.9345390539968907e-05,
+      "loss": 0.6245,
+      "step": 1835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.455047975778258,
+      "learning_rate": 1.934449608436376e-05,
+      "loss": 0.6904,
+      "step": 1836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3158179533536536,
+      "learning_rate": 1.934360103879236e-05,
+      "loss": 0.6602,
+      "step": 1837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1974933152619458,
+      "learning_rate": 1.934270540331122e-05,
+      "loss": 0.6181,
+      "step": 1838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.404012086134293,
+      "learning_rate": 1.934180917797689e-05,
+      "loss": 0.5962,
+      "step": 1839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2609085930556203,
+      "learning_rate": 1.9340912362845946e-05,
+      "loss": 0.613,
+      "step": 1840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2277558953841405,
+      "learning_rate": 1.934001495797501e-05,
+      "loss": 0.6423,
+      "step": 1841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.369381914630601,
+      "learning_rate": 1.9339116963420746e-05,
+      "loss": 0.6398,
+      "step": 1842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.249633306123298,
+      "learning_rate": 1.9338218379239842e-05,
+      "loss": 0.6147,
+      "step": 1843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3382573202992287,
+      "learning_rate": 1.9337319205489035e-05,
+      "loss": 0.6244,
+      "step": 1844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1796217135850182,
+      "learning_rate": 1.933641944222509e-05,
+      "loss": 0.6093,
+      "step": 1845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.260914927378649,
+      "learning_rate": 1.9335519089504816e-05,
+      "loss": 0.5725,
+      "step": 1846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4294441724818887,
+      "learning_rate": 1.933461814738506e-05,
+      "loss": 0.7315,
+      "step": 1847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.607083587461845,
+      "learning_rate": 1.9333716615922696e-05,
+      "loss": 0.6928,
+      "step": 1848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2153931874773367,
+      "learning_rate": 1.933281449517465e-05,
+      "loss": 0.5779,
+      "step": 1849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3373316578675485,
+      "learning_rate": 1.933191178519787e-05,
+      "loss": 0.7226,
+      "step": 1850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.370734925878231,
+      "learning_rate": 1.9331008486049355e-05,
+      "loss": 0.684,
+      "step": 1851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4186321629729854,
+      "learning_rate": 1.933010459778613e-05,
+      "loss": 0.6748,
+      "step": 1852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.187836047357897,
+      "learning_rate": 1.9329200120465268e-05,
+      "loss": 0.6663,
+      "step": 1853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2605960445076303,
+      "learning_rate": 1.932829505414387e-05,
+      "loss": 0.552,
+      "step": 1854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.337893476274334,
+      "learning_rate": 1.9327389398879078e-05,
+      "loss": 0.6848,
+      "step": 1855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4702630062063125,
+      "learning_rate": 1.9326483154728066e-05,
+      "loss": 0.676,
+      "step": 1856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.341186162348423,
+      "learning_rate": 1.9325576321748058e-05,
+      "loss": 0.7112,
+      "step": 1857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2551070313578359,
+      "learning_rate": 1.9324668899996305e-05,
+      "loss": 0.6449,
+      "step": 1858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.140458708882373,
+      "learning_rate": 1.932376088953009e-05,
+      "loss": 0.6111,
+      "step": 1859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2644659315703413,
+      "learning_rate": 1.932285229040675e-05,
+      "loss": 0.631,
+      "step": 1860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.348570450011457,
+      "learning_rate": 1.9321943102683642e-05,
+      "loss": 0.6777,
+      "step": 1861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4518527131950292,
+      "learning_rate": 1.9321033326418174e-05,
+      "loss": 0.6513,
+      "step": 1862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.328931451431817,
+      "learning_rate": 1.932012296166778e-05,
+      "loss": 0.6438,
+      "step": 1863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4358748495985703,
+      "learning_rate": 1.931921200848994e-05,
+      "loss": 0.6652,
+      "step": 1864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3893122531108628,
+      "learning_rate": 1.931830046694216e-05,
+      "loss": 0.6307,
+      "step": 1865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.244040492652964,
+      "learning_rate": 1.9317388337082002e-05,
+      "loss": 0.5596,
+      "step": 1866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2629449513505753,
+      "learning_rate": 1.9316475618967045e-05,
+      "loss": 0.677,
+      "step": 1867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2777468077504603,
+      "learning_rate": 1.9315562312654912e-05,
+      "loss": 0.643,
+      "step": 1868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.381660714514366,
+      "learning_rate": 1.931464841820327e-05,
+      "loss": 0.6704,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1949993546137005,
+      "learning_rate": 1.9313733935669817e-05,
+      "loss": 0.5983,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3454682653193897,
+      "learning_rate": 1.9312818865112288e-05,
+      "loss": 0.6332,
+      "step": 1871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2397029188585333,
+      "learning_rate": 1.9311903206588455e-05,
+      "loss": 0.5938,
+      "step": 1872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2834543014255149,
+      "learning_rate": 1.9310986960156132e-05,
+      "loss": 0.5874,
+      "step": 1873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.257630945217229,
+      "learning_rate": 1.931007012587316e-05,
+      "loss": 0.5996,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.409777772787156,
+      "learning_rate": 1.930915270379743e-05,
+      "loss": 0.6249,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3882650001961634,
+      "learning_rate": 1.930823469398686e-05,
+      "loss": 0.7336,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3206648948843052,
+      "learning_rate": 1.9307316096499412e-05,
+      "loss": 0.6249,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2694585221806376,
+      "learning_rate": 1.9306396911393076e-05,
+      "loss": 0.5818,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3495558467247493,
+      "learning_rate": 1.930547713872589e-05,
+      "loss": 0.6692,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1109811143857988,
+      "learning_rate": 1.930455677855592e-05,
+      "loss": 0.6271,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3275951001573685,
+      "learning_rate": 1.930363583094128e-05,
+      "loss": 0.6553,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4158037119179134,
+      "learning_rate": 1.9302714295940107e-05,
+      "loss": 0.638,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2703026408696196,
+      "learning_rate": 1.9301792173610584e-05,
+      "loss": 0.6525,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.331144468396001,
+      "learning_rate": 1.930086946401093e-05,
+      "loss": 0.628,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1939521977809402,
+      "learning_rate": 1.9299946167199405e-05,
+      "loss": 0.6167,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2909777198570933,
+      "learning_rate": 1.9299022283234288e-05,
+      "loss": 0.6617,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3197922895029108,
+      "learning_rate": 1.9298097812173926e-05,
+      "loss": 0.6428,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.254458439536501,
+      "learning_rate": 1.9297172754076677e-05,
+      "loss": 0.5862,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.264577879833715,
+      "learning_rate": 1.929624710900094e-05,
+      "loss": 0.6829,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2316664422327095,
+      "learning_rate": 1.9295320877005163e-05,
+      "loss": 0.62,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3698721984088331,
+      "learning_rate": 1.9294394058147822e-05,
+      "loss": 0.6855,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2304175956628696,
+      "learning_rate": 1.929346665248743e-05,
+      "loss": 0.5959,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3053276638678097,
+      "learning_rate": 1.9292538660082543e-05,
+      "loss": 0.6279,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1651293071318936,
+      "learning_rate": 1.9291610080991743e-05,
+      "loss": 0.6432,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2475735956473255,
+      "learning_rate": 1.9290680915273662e-05,
+      "loss": 0.6422,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.279957421682411,
+      "learning_rate": 1.928975116298696e-05,
+      "loss": 0.637,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3006048226063103,
+      "learning_rate": 1.928882082419034e-05,
+      "loss": 0.6616,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.244368791243786,
+      "learning_rate": 1.9287889898942537e-05,
+      "loss": 0.7127,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.269686363721072,
+      "learning_rate": 1.9286958387302327e-05,
+      "loss": 0.61,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3600335115623554,
+      "learning_rate": 1.928602628932852e-05,
+      "loss": 0.6265,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2977633363747276,
+      "learning_rate": 1.928509360507996e-05,
+      "loss": 0.6553,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.243239383258666,
+      "learning_rate": 1.928416033461554e-05,
+      "loss": 0.6289,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2940684364328232,
+      "learning_rate": 1.9283226477994173e-05,
+      "loss": 0.5865,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2638571366081577,
+      "learning_rate": 1.928229203527483e-05,
+      "loss": 0.6266,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2164493756498795,
+      "learning_rate": 1.9281357006516496e-05,
+      "loss": 0.6113,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3877431175378758,
+      "learning_rate": 1.928042139177821e-05,
+      "loss": 0.7048,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2450651988227195,
+      "learning_rate": 1.9279485191119042e-05,
+      "loss": 0.58,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.356594574572605,
+      "learning_rate": 1.9278548404598094e-05,
+      "loss": 0.646,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3247773559216494,
+      "learning_rate": 1.9277611032274517e-05,
+      "loss": 0.6741,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3930549821676028,
+      "learning_rate": 1.927667307420749e-05,
+      "loss": 0.6613,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2741597941387712,
+      "learning_rate": 1.927573453045623e-05,
+      "loss": 0.6702,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3242502124390838,
+      "learning_rate": 1.9274795401079987e-05,
+      "loss": 0.6474,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3227753551232575,
+      "learning_rate": 1.9273855686138065e-05,
+      "loss": 0.6436,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2362758149324338,
+      "learning_rate": 1.927291538568979e-05,
+      "loss": 0.6379,
+      "step": 1914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2282747114332857,
+      "learning_rate": 1.9271974499794518e-05,
+      "loss": 0.6129,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2991242264835519,
+      "learning_rate": 1.927103302851166e-05,
+      "loss": 0.6573,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.257815959285931,
+      "learning_rate": 1.9270090971900653e-05,
+      "loss": 0.6114,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.321936021122329,
+      "learning_rate": 1.926914833002098e-05,
+      "loss": 0.6245,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3837403024335373,
+      "learning_rate": 1.9268205102932152e-05,
+      "loss": 0.6319,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.297540562828054,
+      "learning_rate": 1.9267261290693712e-05,
+      "loss": 0.6771,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.425844708428341,
+      "learning_rate": 1.926631689336526e-05,
+      "loss": 0.6402,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2936495460911377,
+      "learning_rate": 1.926537191100641e-05,
+      "loss": 0.6137,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2040072587481967,
+      "learning_rate": 1.9264426343676828e-05,
+      "loss": 0.6284,
+      "step": 1923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1990212144223669,
+      "learning_rate": 1.9263480191436214e-05,
+      "loss": 0.5934,
+      "step": 1924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2670900789528925,
+      "learning_rate": 1.9262533454344303e-05,
+      "loss": 0.6694,
+      "step": 1925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2493116867869654,
+      "learning_rate": 1.9261586132460866e-05,
+      "loss": 0.5705,
+      "step": 1926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3729645228209606,
+      "learning_rate": 1.9260638225845713e-05,
+      "loss": 0.5774,
+      "step": 1927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2335169260311833,
+      "learning_rate": 1.925968973455869e-05,
+      "loss": 0.5917,
+      "step": 1928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2142230346274214,
+      "learning_rate": 1.9258740658659683e-05,
+      "loss": 0.6177,
+      "step": 1929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3049324856390505,
+      "learning_rate": 1.9257790998208606e-05,
+      "loss": 0.697,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.321748392882289,
+      "learning_rate": 1.925684075326542e-05,
+      "loss": 0.6988,
+      "step": 1931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2409835835307796,
+      "learning_rate": 1.9255889923890118e-05,
+      "loss": 0.5907,
+      "step": 1932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2417434282491842,
+      "learning_rate": 1.9254938510142735e-05,
+      "loss": 0.652,
+      "step": 1933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2861551038103585,
+      "learning_rate": 1.925398651208333e-05,
+      "loss": 0.6211,
+      "step": 1934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4243427701804547,
+      "learning_rate": 1.925303392977201e-05,
+      "loss": 0.6335,
+      "step": 1935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.464250286683883,
+      "learning_rate": 1.9252080763268924e-05,
+      "loss": 0.7058,
+      "step": 1936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2647075860172492,
+      "learning_rate": 1.9251127012634242e-05,
+      "loss": 0.6536,
+      "step": 1937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2409793088388401,
+      "learning_rate": 1.9250172677928184e-05,
+      "loss": 0.5969,
+      "step": 1938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3630910046591633,
+      "learning_rate": 1.9249217759211e-05,
+      "loss": 0.6866,
+      "step": 1939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.155029529094975,
+      "learning_rate": 1.924826225654298e-05,
+      "loss": 0.636,
+      "step": 1940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.228628374757012,
+      "learning_rate": 1.9247306169984446e-05,
+      "loss": 0.5917,
+      "step": 1941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3851029989868728,
+      "learning_rate": 1.9246349499595767e-05,
+      "loss": 0.7361,
+      "step": 1942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1385088062808584,
+      "learning_rate": 1.9245392245437336e-05,
+      "loss": 0.618,
+      "step": 1943
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2306579444392938,
+      "learning_rate": 1.9244434407569596e-05,
+      "loss": 0.5768,
+      "step": 1944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2922190541262908,
+      "learning_rate": 1.9243475986053014e-05,
+      "loss": 0.6444,
+      "step": 1945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1461193536916188,
+      "learning_rate": 1.9242516980948105e-05,
+      "loss": 0.5994,
+      "step": 1946
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2921450200876514,
+      "learning_rate": 1.9241557392315413e-05,
+      "loss": 0.6232,
+      "step": 1947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3033486537976007,
+      "learning_rate": 1.9240597220215524e-05,
+      "loss": 0.5464,
+      "step": 1948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2579126258946092,
+      "learning_rate": 1.9239636464709054e-05,
+      "loss": 0.6244,
+      "step": 1949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.269045458061357,
+      "learning_rate": 1.9238675125856666e-05,
+      "loss": 0.6237,
+      "step": 1950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2859785701455273,
+      "learning_rate": 1.923771320371905e-05,
+      "loss": 0.6317,
+      "step": 1951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1734743013465252,
+      "learning_rate": 1.9236750698356942e-05,
+      "loss": 0.6007,
+      "step": 1952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.301467531180983,
+      "learning_rate": 1.9235787609831105e-05,
+      "loss": 0.7346,
+      "step": 1953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2911615563353838,
+      "learning_rate": 1.9234823938202346e-05,
+      "loss": 0.6401,
+      "step": 1954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3560117573833217,
+      "learning_rate": 1.923385968353151e-05,
+      "loss": 0.6993,
+      "step": 1955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2909150192604506,
+      "learning_rate": 1.9232894845879465e-05,
+      "loss": 0.5995,
+      "step": 1956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3388528969166906,
+      "learning_rate": 1.9231929425307136e-05,
+      "loss": 0.7416,
+      "step": 1957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.231763854487558,
+      "learning_rate": 1.9230963421875474e-05,
+      "loss": 0.6337,
+      "step": 1958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2729127422882038,
+      "learning_rate": 1.9229996835645463e-05,
+      "loss": 0.691,
+      "step": 1959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.305057633029852,
+      "learning_rate": 1.9229029666678133e-05,
+      "loss": 0.6141,
+      "step": 1960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3600797470637918,
+      "learning_rate": 1.922806191503454e-05,
+      "loss": 0.6625,
+      "step": 1961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3920475402420074,
+      "learning_rate": 1.9227093580775796e-05,
+      "loss": 0.6639,
+      "step": 1962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2714037891639316,
+      "learning_rate": 1.9226124663963023e-05,
+      "loss": 0.6009,
+      "step": 1963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1792031613870424,
+      "learning_rate": 1.92251551646574e-05,
+      "loss": 0.6305,
+      "step": 1964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2501618757337087,
+      "learning_rate": 1.9224185082920138e-05,
+      "loss": 0.6193,
+      "step": 1965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1921083976531102,
+      "learning_rate": 1.922321441881248e-05,
+      "loss": 0.5905,
+      "step": 1966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3003995409791924,
+      "learning_rate": 1.9222243172395706e-05,
+      "loss": 0.6846,
+      "step": 1967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3237465435699782,
+      "learning_rate": 1.9221271343731146e-05,
+      "loss": 0.6144,
+      "step": 1968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2862052926022414,
+      "learning_rate": 1.922029893288015e-05,
+      "loss": 0.5887,
+      "step": 1969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2031862813031797,
+      "learning_rate": 1.921932593990411e-05,
+      "loss": 0.5889,
+      "step": 1970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4197782210312542,
+      "learning_rate": 1.9218352364864457e-05,
+      "loss": 0.7039,
+      "step": 1971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2854706168123515,
+      "learning_rate": 1.921737820782266e-05,
+      "loss": 0.6811,
+      "step": 1972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.214310507561309,
+      "learning_rate": 1.921640346884022e-05,
+      "loss": 0.5814,
+      "step": 1973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3339426416708104,
+      "learning_rate": 1.9215428147978684e-05,
+      "loss": 0.7675,
+      "step": 1974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2380011208580008,
+      "learning_rate": 1.9214452245299616e-05,
+      "loss": 0.609,
+      "step": 1975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.35708987340541,
+      "learning_rate": 1.9213475760864645e-05,
+      "loss": 0.6503,
+      "step": 1976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1958519210877196,
+      "learning_rate": 1.9212498694735408e-05,
+      "loss": 0.5885,
+      "step": 1977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1480532671635908,
+      "learning_rate": 1.92115210469736e-05,
+      "loss": 0.6093,
+      "step": 1978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2542909406428344,
+      "learning_rate": 1.9210542817640945e-05,
+      "loss": 0.6319,
+      "step": 1979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3220536528360614,
+      "learning_rate": 1.92095640067992e-05,
+      "loss": 0.6659,
+      "step": 1980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4459513309197338,
+      "learning_rate": 1.9208584614510163e-05,
+      "loss": 0.6722,
+      "step": 1981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.264629349141053,
+      "learning_rate": 1.9207604640835675e-05,
+      "loss": 0.623,
+      "step": 1982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.223744989293315,
+      "learning_rate": 1.9206624085837593e-05,
+      "loss": 0.6286,
+      "step": 1983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2420079803302555,
+      "learning_rate": 1.9205642949577835e-05,
+      "loss": 0.6352,
+      "step": 1984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3823857484070743,
+      "learning_rate": 1.9204661232118343e-05,
+      "loss": 0.6804,
+      "step": 1985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2243581997944102,
+      "learning_rate": 1.92036789335211e-05,
+      "loss": 0.5816,
+      "step": 1986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3132825062851852,
+      "learning_rate": 1.920269605384812e-05,
+      "loss": 0.6605,
+      "step": 1987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3311862894492688,
+      "learning_rate": 1.9201712593161458e-05,
+      "loss": 0.641,
+      "step": 1988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3230289301459177,
+      "learning_rate": 1.9200728551523204e-05,
+      "loss": 0.6296,
+      "step": 1989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2198748654931182,
+      "learning_rate": 1.919974392899549e-05,
+      "loss": 0.5966,
+      "step": 1990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2078477388006041,
+      "learning_rate": 1.9198758725640476e-05,
+      "loss": 0.6012,
+      "step": 1991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3875581162112265,
+      "learning_rate": 1.9197772941520365e-05,
+      "loss": 0.6671,
+      "step": 1992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3004054995986967,
+      "learning_rate": 1.9196786576697392e-05,
+      "loss": 0.6817,
+      "step": 1993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3269488793036517,
+      "learning_rate": 1.9195799631233835e-05,
+      "loss": 0.6484,
+      "step": 1994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4012339246169776,
+      "learning_rate": 1.9194812105192003e-05,
+      "loss": 0.7139,
+      "step": 1995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1605691913422316,
+      "learning_rate": 1.9193823998634242e-05,
+      "loss": 0.5975,
+      "step": 1996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.467215690301346,
+      "learning_rate": 1.919283531162294e-05,
+      "loss": 0.7013,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.324162844937865,
+      "learning_rate": 1.9191846044220514e-05,
+      "loss": 0.6951,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2823524384258291,
+      "learning_rate": 1.9190856196489424e-05,
+      "loss": 0.6955,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2663233208022417,
+      "learning_rate": 1.9189865768492168e-05,
+      "loss": 0.5796,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.196024065834504,
+      "learning_rate": 1.918887476029127e-05,
+      "loss": 0.5946,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3400009997207556,
+      "learning_rate": 1.9187883171949298e-05,
+      "loss": 0.6037,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.33373118465062,
+      "learning_rate": 1.9186891003528857e-05,
+      "loss": 0.6111,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.121303684901418,
+      "learning_rate": 1.918589825509259e-05,
+      "loss": 0.5529,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2073844713818416,
+      "learning_rate": 1.9184904926703176e-05,
+      "loss": 0.638,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3220809739976749,
+      "learning_rate": 1.9183911018423324e-05,
+      "loss": 0.6488,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3576937835383587,
+      "learning_rate": 1.9182916530315788e-05,
+      "loss": 0.64,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1506137599116333,
+      "learning_rate": 1.9181921462443354e-05,
+      "loss": 0.6083,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.170778397236292,
+      "learning_rate": 1.9180925814868843e-05,
+      "loss": 0.5477,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3602206789518763,
+      "learning_rate": 1.917992958765512e-05,
+      "loss": 0.6149,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1666669959113247,
+      "learning_rate": 1.917893278086508e-05,
+      "loss": 0.5715,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4121472787584541,
+      "learning_rate": 1.9177935394561652e-05,
+      "loss": 0.6676,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2042105100277287,
+      "learning_rate": 1.9176937428807818e-05,
+      "loss": 0.5866,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2997939570075012,
+      "learning_rate": 1.9175938883666574e-05,
+      "loss": 0.634,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.219849457364017,
+      "learning_rate": 1.9174939759200965e-05,
+      "loss": 0.6235,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2043239020062682,
+      "learning_rate": 1.9173940055474074e-05,
+      "loss": 0.6389,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3267832991276054,
+      "learning_rate": 1.9172939772549014e-05,
+      "loss": 0.634,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3324048365809513,
+      "learning_rate": 1.9171938910488945e-05,
+      "loss": 0.6531,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.358209746341622,
+      "learning_rate": 1.917093746935705e-05,
+      "loss": 0.6804,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3244731287406715,
+      "learning_rate": 1.916993544921655e-05,
+      "loss": 0.6695,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4140043299391518,
+      "learning_rate": 1.9168932850130723e-05,
+      "loss": 0.6657,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2142736931213822,
+      "learning_rate": 1.9167929672162856e-05,
+      "loss": 0.604,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2253866208381643,
+      "learning_rate": 1.9166925915376288e-05,
+      "loss": 0.6437,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3530651405256198,
+      "learning_rate": 1.9165921579834395e-05,
+      "loss": 0.6696,
+      "step": 2024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2925642596711908,
+      "learning_rate": 1.916491666560058e-05,
+      "loss": 0.6431,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5131163925310542,
+      "learning_rate": 1.916391117273829e-05,
+      "loss": 0.6546,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.270793674454692,
+      "learning_rate": 1.916290510131101e-05,
+      "loss": 0.6366,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2982945314141887,
+      "learning_rate": 1.9161898451382257e-05,
+      "loss": 0.6531,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3241880970335667,
+      "learning_rate": 1.9160891223015586e-05,
+      "loss": 0.6709,
+      "step": 2029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2966803439104615,
+      "learning_rate": 1.9159883416274585e-05,
+      "loss": 0.6174,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.174956326991961,
+      "learning_rate": 1.915887503122289e-05,
+      "loss": 0.604,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3852026590059048,
+      "learning_rate": 1.9157866067924157e-05,
+      "loss": 0.6698,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.341433768859375,
+      "learning_rate": 1.9156856526442092e-05,
+      "loss": 0.6054,
+      "step": 2033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3624354583432348,
+      "learning_rate": 1.915584640684043e-05,
+      "loss": 0.6264,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2086475007742867,
+      "learning_rate": 1.9154835709182947e-05,
+      "loss": 0.6339,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3800078576320352,
+      "learning_rate": 1.9153824433533453e-05,
+      "loss": 0.622,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2407871246102744,
+      "learning_rate": 1.9152812579955795e-05,
+      "loss": 0.6017,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3450462831099477,
+      "learning_rate": 1.915180014851386e-05,
+      "loss": 0.6193,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3559467890857104,
+      "learning_rate": 1.915078713927156e-05,
+      "loss": 0.6751,
+      "step": 2039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2963046922613652,
+      "learning_rate": 1.9149773552292855e-05,
+      "loss": 0.631,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2528847309470341,
+      "learning_rate": 1.9148759387641745e-05,
+      "loss": 0.554,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1992937493247826,
+      "learning_rate": 1.914774464538225e-05,
+      "loss": 0.5803,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3610252360849864,
+      "learning_rate": 1.914672932557844e-05,
+      "loss": 0.6453,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4820835295340933,
+      "learning_rate": 1.9145713428294415e-05,
+      "loss": 0.7304,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.37786458467916,
+      "learning_rate": 1.9144696953594316e-05,
+      "loss": 0.7114,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.207204565549222,
+      "learning_rate": 1.9143679901542316e-05,
+      "loss": 0.6328,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4232089314528413,
+      "learning_rate": 1.914266227220263e-05,
+      "loss": 0.6826,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3197123952581942,
+      "learning_rate": 1.9141644065639507e-05,
+      "loss": 0.655,
+      "step": 2048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2843356309783234,
+      "learning_rate": 1.914062528191723e-05,
+      "loss": 0.6741,
+      "step": 2049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.209933646250681,
+      "learning_rate": 1.9139605921100116e-05,
+      "loss": 0.6104,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2265959814082021,
+      "learning_rate": 1.9138585983252527e-05,
+      "loss": 0.6151,
+      "step": 2051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.243258128836326,
+      "learning_rate": 1.913756546843886e-05,
+      "loss": 0.5764,
+      "step": 2052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.288175104046891,
+      "learning_rate": 1.9136544376723537e-05,
+      "loss": 0.6096,
+      "step": 2053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2952533832675945,
+      "learning_rate": 1.9135522708171034e-05,
+      "loss": 0.6438,
+      "step": 2054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2134435010943878,
+      "learning_rate": 1.9134500462845844e-05,
+      "loss": 0.5902,
+      "step": 2055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2717059948707234,
+      "learning_rate": 1.9133477640812513e-05,
+      "loss": 0.5554,
+      "step": 2056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3004287379538864,
+      "learning_rate": 1.9132454242135618e-05,
+      "loss": 0.6017,
+      "step": 2057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2478443154497603,
+      "learning_rate": 1.9131430266879766e-05,
+      "loss": 0.5813,
+      "step": 2058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.451664180089235,
+      "learning_rate": 1.913040571510961e-05,
+      "loss": 0.6311,
+      "step": 2059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2436437647702974,
+      "learning_rate": 1.9129380586889836e-05,
+      "loss": 0.585,
+      "step": 2060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.231593365886224,
+      "learning_rate": 1.9128354882285166e-05,
+      "loss": 0.6128,
+      "step": 2061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1896508965398727,
+      "learning_rate": 1.9127328601360354e-05,
+      "loss": 0.5924,
+      "step": 2062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3441368255960633,
+      "learning_rate": 1.91263017441802e-05,
+      "loss": 0.67,
+      "step": 2063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.273881799619276,
+      "learning_rate": 1.9125274310809524e-05,
+      "loss": 0.6737,
+      "step": 2064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2376626572493064,
+      "learning_rate": 1.9124246301313206e-05,
+      "loss": 0.584,
+      "step": 2065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.366355555680691,
+      "learning_rate": 1.9123217715756142e-05,
+      "loss": 0.6793,
+      "step": 2066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3335146879997093,
+      "learning_rate": 1.9122188554203275e-05,
+      "loss": 0.6198,
+      "step": 2067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3164377180464333,
+      "learning_rate": 1.9121158816719577e-05,
+      "loss": 0.6492,
+      "step": 2068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3035112663838624,
+      "learning_rate": 1.912012850337007e-05,
+      "loss": 0.5759,
+      "step": 2069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4558794720825778,
+      "learning_rate": 1.9119097614219795e-05,
+      "loss": 0.6725,
+      "step": 2070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.202374707394164,
+      "learning_rate": 1.911806614933384e-05,
+      "loss": 0.6078,
+      "step": 2071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.217059111771495,
+      "learning_rate": 1.9117034108777323e-05,
+      "loss": 0.6125,
+      "step": 2072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2498669553525101,
+      "learning_rate": 1.9116001492615403e-05,
+      "loss": 0.6607,
+      "step": 2073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.425125932984774,
+      "learning_rate": 1.911496830091328e-05,
+      "loss": 0.7054,
+      "step": 2074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.273080834365659,
+      "learning_rate": 1.911393453373618e-05,
+      "loss": 0.6738,
+      "step": 2075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3463576985158532,
+      "learning_rate": 1.9112900191149374e-05,
+      "loss": 0.6432,
+      "step": 2076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3587925419468845,
+      "learning_rate": 1.911186527321816e-05,
+      "loss": 0.6974,
+      "step": 2077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3472790618695383,
+      "learning_rate": 1.911082978000788e-05,
+      "loss": 0.5873,
+      "step": 2078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2716541089095745,
+      "learning_rate": 1.910979371158391e-05,
+      "loss": 0.6436,
+      "step": 2079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3235672780148966,
+      "learning_rate": 1.910875706801166e-05,
+      "loss": 0.601,
+      "step": 2080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3177046895133842,
+      "learning_rate": 1.9107719849356588e-05,
+      "loss": 0.6769,
+      "step": 2081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.272032635641989,
+      "learning_rate": 1.9106682055684168e-05,
+      "loss": 0.6449,
+      "step": 2082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.250553247089158,
+      "learning_rate": 1.9105643687059926e-05,
+      "loss": 0.6294,
+      "step": 2083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1390032801258525,
+      "learning_rate": 1.9104604743549422e-05,
+      "loss": 0.5746,
+      "step": 2084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1694509739868435,
+      "learning_rate": 1.9103565225218243e-05,
+      "loss": 0.6365,
+      "step": 2085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2251363834054254,
+      "learning_rate": 1.9102525132132028e-05,
+      "loss": 0.6455,
+      "step": 2086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3180199758436981,
+      "learning_rate": 1.910148446435643e-05,
+      "loss": 0.6392,
+      "step": 2087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3194633778111564,
+      "learning_rate": 1.910044322195717e-05,
+      "loss": 0.562,
+      "step": 2088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.288310299667858,
+      "learning_rate": 1.9099401404999972e-05,
+      "loss": 0.6323,
+      "step": 2089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3034362732235119,
+      "learning_rate": 1.9098359013550617e-05,
+      "loss": 0.6591,
+      "step": 2090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2375088758824635,
+      "learning_rate": 1.9097316047674915e-05,
+      "loss": 0.6036,
+      "step": 2091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2155822275673247,
+      "learning_rate": 1.9096272507438715e-05,
+      "loss": 0.6065,
+      "step": 2092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3515791699037465,
+      "learning_rate": 1.9095228392907904e-05,
+      "loss": 0.6858,
+      "step": 2093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2961803552206335,
+      "learning_rate": 1.90941837041484e-05,
+      "loss": 0.669,
+      "step": 2094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2780059591133508,
+      "learning_rate": 1.9093138441226156e-05,
+      "loss": 0.5898,
+      "step": 2095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8588883861136467,
+      "learning_rate": 1.9092092604207166e-05,
+      "loss": 0.7157,
+      "step": 2096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.175396675761132,
+      "learning_rate": 1.9091046193157464e-05,
+      "loss": 0.6107,
+      "step": 2097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3872467325364235,
+      "learning_rate": 1.9089999208143113e-05,
+      "loss": 0.6484,
+      "step": 2098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3398876127397998,
+      "learning_rate": 1.908895164923021e-05,
+      "loss": 0.6737,
+      "step": 2099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3554141338679262,
+      "learning_rate": 1.9087903516484898e-05,
+      "loss": 0.6957,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2921149901188436,
+      "learning_rate": 1.908685480997335e-05,
+      "loss": 0.6873,
+      "step": 2101
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6798620088990228,
+      "learning_rate": 1.9085805529761778e-05,
+      "loss": 0.6263,
+      "step": 2102
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1164184804557442,
+      "learning_rate": 1.9084755675916423e-05,
+      "loss": 0.5341,
+      "step": 2103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.306438937965065,
+      "learning_rate": 1.9083705248503575e-05,
+      "loss": 0.6658,
+      "step": 2104
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1967670840224751,
+      "learning_rate": 1.9082654247589543e-05,
+      "loss": 0.6296,
+      "step": 2105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3573343591814218,
+      "learning_rate": 1.9081602673240695e-05,
+      "loss": 0.6393,
+      "step": 2106
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2865097663677707,
+      "learning_rate": 1.9080550525523413e-05,
+      "loss": 0.6403,
+      "step": 2107
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1711245868025382,
+      "learning_rate": 1.9079497804504122e-05,
+      "loss": 0.5849,
+      "step": 2108
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3722672616377143,
+      "learning_rate": 1.9078444510249294e-05,
+      "loss": 0.6448,
+      "step": 2109
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3890895725010728,
+      "learning_rate": 1.9077390642825427e-05,
+      "loss": 0.692,
+      "step": 2110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2972426295575574,
+      "learning_rate": 1.9076336202299055e-05,
+      "loss": 0.6433,
+      "step": 2111
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.245771073324085,
+      "learning_rate": 1.907528118873675e-05,
+      "loss": 0.6449,
+      "step": 2112
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2418793589991117,
+      "learning_rate": 1.907422560220512e-05,
+      "loss": 0.6897,
+      "step": 2113
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.222195767106704,
+      "learning_rate": 1.9073169442770814e-05,
+      "loss": 0.6081,
+      "step": 2114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2152395309000437,
+      "learning_rate": 1.907211271050051e-05,
+      "loss": 0.6626,
+      "step": 2115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.26159558713706,
+      "learning_rate": 1.907105540546092e-05,
+      "loss": 0.69,
+      "step": 2116
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2454860245675936,
+      "learning_rate": 1.9069997527718803e-05,
+      "loss": 0.6623,
+      "step": 2117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.206306956236214,
+      "learning_rate": 1.906893907734095e-05,
+      "loss": 0.6518,
+      "step": 2118
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2104302605255737,
+      "learning_rate": 1.9067880054394182e-05,
+      "loss": 0.6932,
+      "step": 2119
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1725419753151858,
+      "learning_rate": 1.906682045894536e-05,
+      "loss": 0.6345,
+      "step": 2120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3535865666017106,
+      "learning_rate": 1.9065760291061385e-05,
+      "loss": 0.6849,
+      "step": 2121
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2504353718734038,
+      "learning_rate": 1.9064699550809193e-05,
+      "loss": 0.6639,
+      "step": 2122
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2486380309806338,
+      "learning_rate": 1.9063638238255747e-05,
+      "loss": 0.6013,
+      "step": 2123
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2627985440780807,
+      "learning_rate": 1.9062576353468055e-05,
+      "loss": 0.6623,
+      "step": 2124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2502061197093575,
+      "learning_rate": 1.9061513896513166e-05,
+      "loss": 0.6208,
+      "step": 2125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5639339733382052,
+      "learning_rate": 1.9060450867458152e-05,
+      "loss": 0.7525,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2891186326260917,
+      "learning_rate": 1.9059387266370127e-05,
+      "loss": 0.6546,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.555948484082821,
+      "learning_rate": 1.9058323093316247e-05,
+      "loss": 0.6851,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2131105675052543,
+      "learning_rate": 1.905725834836369e-05,
+      "loss": 0.6762,
+      "step": 2129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3119491147663953,
+      "learning_rate": 1.905619303157969e-05,
+      "loss": 0.6108,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3226456652430132,
+      "learning_rate": 1.9055127143031495e-05,
+      "loss": 0.6548,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3480355876051007,
+      "learning_rate": 1.905406068278641e-05,
+      "loss": 0.7146,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3381445427954821,
+      "learning_rate": 1.9052993650911758e-05,
+      "loss": 0.6786,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2937735274382405,
+      "learning_rate": 1.905192604747491e-05,
+      "loss": 0.6446,
+      "step": 2134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2406057207353547,
+      "learning_rate": 1.905085787254327e-05,
+      "loss": 0.6253,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2631125295590657,
+      "learning_rate": 1.9049789126184275e-05,
+      "loss": 0.6205,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1535828216543516,
+      "learning_rate": 1.90487198084654e-05,
+      "loss": 0.6474,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3129414315623575,
+      "learning_rate": 1.904764991945416e-05,
+      "loss": 0.6226,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1583938827515148,
+      "learning_rate": 1.9046579459218103e-05,
+      "loss": 0.6406,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2867256483994411,
+      "learning_rate": 1.904550842782481e-05,
+      "loss": 0.6571,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2765504199125304,
+      "learning_rate": 1.90444368253419e-05,
+      "loss": 0.64,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2056213836061969,
+      "learning_rate": 1.9043364651837026e-05,
+      "loss": 0.6632,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2662644360426416,
+      "learning_rate": 1.9042291907377886e-05,
+      "loss": 0.674,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2909196826644382,
+      "learning_rate": 1.904121859203221e-05,
+      "loss": 0.6332,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1797502008410858,
+      "learning_rate": 1.9040144705867755e-05,
+      "loss": 0.5879,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0404876457958412,
+      "learning_rate": 1.9039070248952324e-05,
+      "loss": 0.6045,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2474719232063831,
+      "learning_rate": 1.9037995221353754e-05,
+      "loss": 0.6385,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.285709853202506,
+      "learning_rate": 1.9036919623139916e-05,
+      "loss": 0.611,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.22290587390684,
+      "learning_rate": 1.9035843454378715e-05,
+      "loss": 0.5861,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3315507128287138,
+      "learning_rate": 1.90347667151381e-05,
+      "loss": 0.5779,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3597209534050894,
+      "learning_rate": 1.903368940548605e-05,
+      "loss": 0.6834,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1954379047108512,
+      "learning_rate": 1.9032611525490575e-05,
+      "loss": 0.6113,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.270987371062318,
+      "learning_rate": 1.9031533075219737e-05,
+      "loss": 0.6587,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2807978320656799,
+      "learning_rate": 1.903045405474162e-05,
+      "loss": 0.582,
+      "step": 2154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2917413792485852,
+      "learning_rate": 1.9029374464124344e-05,
+      "loss": 0.651,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3435415061839777,
+      "learning_rate": 1.902829430343607e-05,
+      "loss": 0.6278,
+      "step": 2156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1928243733335564,
+      "learning_rate": 1.9027213572745006e-05,
+      "loss": 0.59,
+      "step": 2157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.393836738042806,
+      "learning_rate": 1.9026132272119368e-05,
+      "loss": 0.6737,
+      "step": 2158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3304006938809796,
+      "learning_rate": 1.9025050401627433e-05,
+      "loss": 0.6413,
+      "step": 2159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3030657711336548,
+      "learning_rate": 1.90239679613375e-05,
+      "loss": 0.6559,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2609678697296773,
+      "learning_rate": 1.9022884951317916e-05,
+      "loss": 0.5928,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2555601436864992,
+      "learning_rate": 1.9021801371637055e-05,
+      "loss": 0.6179,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2895793918843141,
+      "learning_rate": 1.9020717222363322e-05,
+      "loss": 0.6373,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2674462671571585,
+      "learning_rate": 1.9019632503565174e-05,
+      "loss": 0.583,
+      "step": 2164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.25094044593771,
+      "learning_rate": 1.9018547215311086e-05,
+      "loss": 0.6169,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1661626317374814,
+      "learning_rate": 1.9017461357669588e-05,
+      "loss": 0.6119,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3096499607292253,
+      "learning_rate": 1.9016374930709223e-05,
+      "loss": 0.6436,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2719579889839332,
+      "learning_rate": 1.9015287934498598e-05,
+      "loss": 0.6431,
+      "step": 2168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3271048274113608,
+      "learning_rate": 1.901420036910633e-05,
+      "loss": 0.6619,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2545936102790916,
+      "learning_rate": 1.9013112234601084e-05,
+      "loss": 0.6566,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3744451964086848,
+      "learning_rate": 1.9012023531051565e-05,
+      "loss": 0.6734,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2590793366678148,
+      "learning_rate": 1.90109342585265e-05,
+      "loss": 0.6081,
+      "step": 2172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2601475809346592,
+      "learning_rate": 1.900984441709467e-05,
+      "loss": 0.6688,
+      "step": 2173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1182569173713621,
+      "learning_rate": 1.9008754006824874e-05,
+      "loss": 0.5623,
+      "step": 2174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3822473351681779,
+      "learning_rate": 1.900766302778596e-05,
+      "loss": 0.6451,
+      "step": 2175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2689384132633361,
+      "learning_rate": 1.9006571480046805e-05,
+      "loss": 0.61,
+      "step": 2176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3104545003567227,
+      "learning_rate": 1.9005479363676322e-05,
+      "loss": 0.6262,
+      "step": 2177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3082689024414194,
+      "learning_rate": 1.9004386678743468e-05,
+      "loss": 0.6298,
+      "step": 2178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3352777737155743,
+      "learning_rate": 1.9003293425317224e-05,
+      "loss": 0.6366,
+      "step": 2179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3554340544995567,
+      "learning_rate": 1.9002199603466617e-05,
+      "loss": 0.6191,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2130912577964308,
+      "learning_rate": 1.90011052132607e-05,
+      "loss": 0.6114,
+      "step": 2181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.311628006596029,
+      "learning_rate": 1.9000010254768576e-05,
+      "loss": 0.6532,
+      "step": 2182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1812256644027996,
+      "learning_rate": 1.899891472805937e-05,
+      "loss": 0.5825,
+      "step": 2183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2512998498214598,
+      "learning_rate": 1.899781863320225e-05,
+      "loss": 0.6545,
+      "step": 2184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2153521881332408,
+      "learning_rate": 1.8996721970266417e-05,
+      "loss": 0.6306,
+      "step": 2185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2840863444123651,
+      "learning_rate": 1.8995624739321104e-05,
+      "loss": 0.6233,
+      "step": 2186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2849655241738704,
+      "learning_rate": 1.8994526940435593e-05,
+      "loss": 0.6324,
+      "step": 2187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2846121984343022,
+      "learning_rate": 1.8993428573679193e-05,
+      "loss": 0.6277,
+      "step": 2188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3546217764956616,
+      "learning_rate": 1.8992329639121243e-05,
+      "loss": 0.5932,
+      "step": 2189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.221647974177459,
+      "learning_rate": 1.8991230136831133e-05,
+      "loss": 0.6211,
+      "step": 2190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2044068480845211,
+      "learning_rate": 1.899013006687827e-05,
+      "loss": 0.6546,
+      "step": 2191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2214997888659351,
+      "learning_rate": 1.8989029429332117e-05,
+      "loss": 0.6488,
+      "step": 2192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.217256805053041,
+      "learning_rate": 1.898792822426216e-05,
+      "loss": 0.6382,
+      "step": 2193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3706500430413289,
+      "learning_rate": 1.8986826451737918e-05,
+      "loss": 0.6574,
+      "step": 2194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3003808399037404,
+      "learning_rate": 1.898572411182896e-05,
+      "loss": 0.7057,
+      "step": 2195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3763865069357932,
+      "learning_rate": 1.898462120460488e-05,
+      "loss": 0.6532,
+      "step": 2196
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3425234586204093,
+      "learning_rate": 1.8983517730135304e-05,
+      "loss": 0.5741,
+      "step": 2197
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2425613320639064,
+      "learning_rate": 1.8982413688489906e-05,
+      "loss": 0.6441,
+      "step": 2198
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2721903492036613,
+      "learning_rate": 1.898130907973839e-05,
+      "loss": 0.6709,
+      "step": 2199
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2513695843207595,
+      "learning_rate": 1.8980203903950495e-05,
+      "loss": 0.6058,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3495815953267751,
+      "learning_rate": 1.8979098161195995e-05,
+      "loss": 0.7001,
+      "step": 2201
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.237918981273691,
+      "learning_rate": 1.89779918515447e-05,
+      "loss": 0.5909,
+      "step": 2202
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.236287241375756,
+      "learning_rate": 1.8976884975066464e-05,
+      "loss": 0.5965,
+      "step": 2203
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.272863996116338,
+      "learning_rate": 1.897577753183116e-05,
+      "loss": 0.6082,
+      "step": 2204
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2897863494282011,
+      "learning_rate": 1.8974669521908714e-05,
+      "loss": 0.6891,
+      "step": 2205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1921505963574597,
+      "learning_rate": 1.8973560945369076e-05,
+      "loss": 0.6242,
+      "step": 2206
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1876158908711067,
+      "learning_rate": 1.897245180228224e-05,
+      "loss": 0.6502,
+      "step": 2207
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2537498495419412,
+      "learning_rate": 1.897134209271823e-05,
+      "loss": 0.6333,
+      "step": 2208
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1121322859938865,
+      "learning_rate": 1.8970231816747103e-05,
+      "loss": 0.5998,
+      "step": 2209
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2086086398035403,
+      "learning_rate": 1.8969120974438967e-05,
+      "loss": 0.5792,
+      "step": 2210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.405265208969785,
+      "learning_rate": 1.896800956586395e-05,
+      "loss": 0.7066,
+      "step": 2211
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2142595069926199,
+      "learning_rate": 1.8966897591092213e-05,
+      "loss": 0.6043,
+      "step": 2212
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3413202808094546,
+      "learning_rate": 1.8965785050193976e-05,
+      "loss": 0.6512,
+      "step": 2213
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.207949144300272,
+      "learning_rate": 1.8964671943239467e-05,
+      "loss": 0.6276,
+      "step": 2214
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1663273533623015,
+      "learning_rate": 1.8963558270298965e-05,
+      "loss": 0.6386,
+      "step": 2215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1964878960239114,
+      "learning_rate": 1.8962444031442788e-05,
+      "loss": 0.6284,
+      "step": 2216
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3188564691457447,
+      "learning_rate": 1.8961329226741277e-05,
+      "loss": 0.6064,
+      "step": 2217
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2804359082705383,
+      "learning_rate": 1.8960213856264818e-05,
+      "loss": 0.6543,
+      "step": 2218
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2107124365572628,
+      "learning_rate": 1.8959097920083828e-05,
+      "loss": 0.5967,
+      "step": 2219
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4385573603358472,
+      "learning_rate": 1.8957981418268764e-05,
+      "loss": 0.6438,
+      "step": 2220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1374573814875977,
+      "learning_rate": 1.8956864350890117e-05,
+      "loss": 0.5845,
+      "step": 2221
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.428789617701718,
+      "learning_rate": 1.8955746718018413e-05,
+      "loss": 0.657,
+      "step": 2222
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2356568453075751,
+      "learning_rate": 1.895462851972421e-05,
+      "loss": 0.6111,
+      "step": 2223
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1890877599451777,
+      "learning_rate": 1.895350975607811e-05,
+      "loss": 0.5742,
+      "step": 2224
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2546147041533051,
+      "learning_rate": 1.8952390427150747e-05,
+      "loss": 0.6602,
+      "step": 2225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.269079227670734,
+      "learning_rate": 1.8951270533012786e-05,
+      "loss": 0.6248,
+      "step": 2226
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3069233723850515,
+      "learning_rate": 1.8950150073734937e-05,
+      "loss": 0.6711,
+      "step": 2227
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3370850302563713,
+      "learning_rate": 1.8949029049387933e-05,
+      "loss": 0.6741,
+      "step": 2228
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3543312510706706,
+      "learning_rate": 1.8947907460042558e-05,
+      "loss": 0.5988,
+      "step": 2229
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3573715529959696,
+      "learning_rate": 1.8946785305769616e-05,
+      "loss": 0.6502,
+      "step": 2230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2871090045285887,
+      "learning_rate": 1.894566258663996e-05,
+      "loss": 0.6669,
+      "step": 2231
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.20930375670513,
+      "learning_rate": 1.894453930272447e-05,
+      "loss": 0.5849,
+      "step": 2232
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2904696558089934,
+      "learning_rate": 1.8943415454094068e-05,
+      "loss": 0.5952,
+      "step": 2233
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3465359218946358,
+      "learning_rate": 1.894229104081971e-05,
+      "loss": 0.6687,
+      "step": 2234
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3892760431259157,
+      "learning_rate": 1.8941166062972374e-05,
+      "loss": 0.6687,
+      "step": 2235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.172477719760186,
+      "learning_rate": 1.89400405206231e-05,
+      "loss": 0.5805,
+      "step": 2236
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.329783089022568,
+      "learning_rate": 1.893891441384294e-05,
+      "loss": 0.6709,
+      "step": 2237
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1965713354797212,
+      "learning_rate": 1.8937787742703e-05,
+      "loss": 0.5919,
+      "step": 2238
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1834011109959846,
+      "learning_rate": 1.8936660507274403e-05,
+      "loss": 0.6151,
+      "step": 2239
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2269391010191004,
+      "learning_rate": 1.8935532707628322e-05,
+      "loss": 0.6235,
+      "step": 2240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.398545969587041,
+      "learning_rate": 1.8934404343835956e-05,
+      "loss": 0.7054,
+      "step": 2241
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2660875299426408,
+      "learning_rate": 1.893327541596855e-05,
+      "loss": 0.675,
+      "step": 2242
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.350481736803621,
+      "learning_rate": 1.893214592409738e-05,
+      "loss": 0.6923,
+      "step": 2243
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.257509230547599,
+      "learning_rate": 1.8931015868293755e-05,
+      "loss": 0.6031,
+      "step": 2244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2712820803365117,
+      "learning_rate": 1.8929885248629017e-05,
+      "loss": 0.6541,
+      "step": 2245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3138848673432098,
+      "learning_rate": 1.8928754065174552e-05,
+      "loss": 0.6442,
+      "step": 2246
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1687879464420263,
+      "learning_rate": 1.8927622318001778e-05,
+      "loss": 0.6104,
+      "step": 2247
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1142975033670908,
+      "learning_rate": 1.8926490007182147e-05,
+      "loss": 0.5627,
+      "step": 2248
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2697622705217075,
+      "learning_rate": 1.8925357132787142e-05,
+      "loss": 0.6007,
+      "step": 2249
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1813173466613234,
+      "learning_rate": 1.8924223694888297e-05,
+      "loss": 0.5944,
+      "step": 2250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5465209247445078,
+      "learning_rate": 1.8923089693557165e-05,
+      "loss": 0.6431,
+      "step": 2251
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2957509696437204,
+      "learning_rate": 1.8921955128865343e-05,
+      "loss": 0.6926,
+      "step": 2252
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2173761302434725,
+      "learning_rate": 1.8920820000884467e-05,
+      "loss": 0.5983,
+      "step": 2253
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.177543156474835,
+      "learning_rate": 1.8919684309686193e-05,
+      "loss": 0.5842,
+      "step": 2254
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.412638542672742,
+      "learning_rate": 1.8918548055342232e-05,
+      "loss": 0.6856,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3146443334211353,
+      "learning_rate": 1.891741123792432e-05,
+      "loss": 0.6706,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2936881561977949,
+      "learning_rate": 1.891627385750423e-05,
+      "loss": 0.6069,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.334859491161905,
+      "learning_rate": 1.8915135914153766e-05,
+      "loss": 0.7092,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3024241751567636,
+      "learning_rate": 1.8913997407944776e-05,
+      "loss": 0.6325,
+      "step": 2259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.251937128164181,
+      "learning_rate": 1.891285833894914e-05,
+      "loss": 0.6681,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2178926631316538,
+      "learning_rate": 1.8911718707238772e-05,
+      "loss": 0.6195,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.358931765150668,
+      "learning_rate": 1.8910578512885624e-05,
+      "loss": 0.6746,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2550699888524264,
+      "learning_rate": 1.8909437755961683e-05,
+      "loss": 0.5919,
+      "step": 2263
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1356355940178202,
+      "learning_rate": 1.890829643653897e-05,
+      "loss": 0.642,
+      "step": 2264
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2172656189673872,
+      "learning_rate": 1.890715455468954e-05,
+      "loss": 0.628,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.258034491052529,
+      "learning_rate": 1.890601211048549e-05,
+      "loss": 0.6401,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.333558937695525,
+      "learning_rate": 1.8904869103998947e-05,
+      "loss": 0.691,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2932833378180306,
+      "learning_rate": 1.8903725535302073e-05,
+      "loss": 0.6411,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2341741084549087,
+      "learning_rate": 1.8902581404467067e-05,
+      "loss": 0.6379,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2396277198003882,
+      "learning_rate": 1.890143671156617e-05,
+      "loss": 0.5427,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2724096912946752,
+      "learning_rate": 1.8900291456671645e-05,
+      "loss": 0.5593,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2147699560585856,
+      "learning_rate": 1.8899145639855803e-05,
+      "loss": 0.6228,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3003643845683441,
+      "learning_rate": 1.889799926119098e-05,
+      "loss": 0.5676,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.385570168655623,
+      "learning_rate": 1.8896852320749558e-05,
+      "loss": 0.6653,
+      "step": 2274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2159537000907827,
+      "learning_rate": 1.8895704818603947e-05,
+      "loss": 0.6018,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.346308468228186,
+      "learning_rate": 1.8894556754826597e-05,
+      "loss": 0.6397,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.331505903980833,
+      "learning_rate": 1.889340812948999e-05,
+      "loss": 0.6712,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.337518080473291,
+      "learning_rate": 1.8892258942666637e-05,
+      "loss": 0.6874,
+      "step": 2278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3073913971853512,
+      "learning_rate": 1.8891109194429102e-05,
+      "loss": 0.6233,
+      "step": 2279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2000252065395383,
+      "learning_rate": 1.8889958884849972e-05,
+      "loss": 0.6291,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2879512273930918,
+      "learning_rate": 1.8888808014001874e-05,
+      "loss": 0.6618,
+      "step": 2281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3074045271872268,
+      "learning_rate": 1.8887656581957464e-05,
+      "loss": 0.7366,
+      "step": 2282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.334403646550239,
+      "learning_rate": 1.888650458878944e-05,
+      "loss": 0.6571,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1870009728984494,
+      "learning_rate": 1.888535203457053e-05,
+      "loss": 0.5928,
+      "step": 2284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2679478087138498,
+      "learning_rate": 1.8884198919373507e-05,
+      "loss": 0.6118,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2155057323953022,
+      "learning_rate": 1.8883045243271168e-05,
+      "loss": 0.6207,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3155058546323852,
+      "learning_rate": 1.8881891006336354e-05,
+      "loss": 0.6429,
+      "step": 2287
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3744120641327664,
+      "learning_rate": 1.8880736208641937e-05,
+      "loss": 0.6492,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2442458750522478,
+      "learning_rate": 1.887958085026082e-05,
+      "loss": 0.6134,
+      "step": 2289
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3569818675489023,
+      "learning_rate": 1.8878424931265957e-05,
+      "loss": 0.6853,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3909016880420917,
+      "learning_rate": 1.887726845173032e-05,
+      "loss": 0.691,
+      "step": 2291
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1964597494621914,
+      "learning_rate": 1.887611141172692e-05,
+      "loss": 0.6377,
+      "step": 2292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2263453102985502,
+      "learning_rate": 1.8874953811328817e-05,
+      "loss": 0.6161,
+      "step": 2293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3458127889255969,
+      "learning_rate": 1.8873795650609092e-05,
+      "loss": 0.7277,
+      "step": 2294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2911352889889798,
+      "learning_rate": 1.8872636929640863e-05,
+      "loss": 0.6348,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3019295970440106,
+      "learning_rate": 1.8871477648497285e-05,
+      "loss": 0.674,
+      "step": 2296
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2017848489944882,
+      "learning_rate": 1.8870317807251555e-05,
+      "loss": 0.6484,
+      "step": 2297
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1973057004310674,
+      "learning_rate": 1.8869157405976896e-05,
+      "loss": 0.6222,
+      "step": 2298
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2511482210332108,
+      "learning_rate": 1.8867996444746574e-05,
+      "loss": 0.598,
+      "step": 2299
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2101373797842334,
+      "learning_rate": 1.886683492363388e-05,
+      "loss": 0.6136,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2807353777254566,
+      "learning_rate": 1.886567284271215e-05,
+      "loss": 0.6396,
+      "step": 2301
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2229585608996965,
+      "learning_rate": 1.886451020205476e-05,
+      "loss": 0.6031,
+      "step": 2302
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4295672590211297,
+      "learning_rate": 1.8863347001735098e-05,
+      "loss": 0.6923,
+      "step": 2303
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.24686252712044,
+      "learning_rate": 1.8862183241826613e-05,
+      "loss": 0.6227,
+      "step": 2304
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2491685009095252,
+      "learning_rate": 1.8861018922402776e-05,
+      "loss": 0.6716,
+      "step": 2305
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.323084342522879,
+      "learning_rate": 1.88598540435371e-05,
+      "loss": 0.6527,
+      "step": 2306
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.317985425118884,
+      "learning_rate": 1.8858688605303127e-05,
+      "loss": 0.6852,
+      "step": 2307
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5015658312074862,
+      "learning_rate": 1.8857522607774438e-05,
+      "loss": 0.6642,
+      "step": 2308
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2061895009419612,
+      "learning_rate": 1.8856356051024646e-05,
+      "loss": 0.6094,
+      "step": 2309
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1317907526773971,
+      "learning_rate": 1.8855188935127405e-05,
+      "loss": 0.5303,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.247687823923802,
+      "learning_rate": 1.88540212601564e-05,
+      "loss": 0.6095,
+      "step": 2311
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2911394899537405,
+      "learning_rate": 1.8852853026185348e-05,
+      "loss": 0.6758,
+      "step": 2312
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2499499310956306,
+      "learning_rate": 1.8851684233288016e-05,
+      "loss": 0.5879,
+      "step": 2313
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4953969423373774,
+      "learning_rate": 1.8850514881538186e-05,
+      "loss": 0.6632,
+      "step": 2314
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1763402174141706,
+      "learning_rate": 1.884934497100969e-05,
+      "loss": 0.5729,
+      "step": 2315
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1899228225653142,
+      "learning_rate": 1.8848174501776388e-05,
+      "loss": 0.6233,
+      "step": 2316
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3122335799344167,
+      "learning_rate": 1.8847003473912182e-05,
+      "loss": 0.5731,
+      "step": 2317
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2591900597416277,
+      "learning_rate": 1.8845831887490998e-05,
+      "loss": 0.6578,
+      "step": 2318
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2976041835051901,
+      "learning_rate": 1.8844659742586813e-05,
+      "loss": 0.6085,
+      "step": 2319
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1819868100845572,
+      "learning_rate": 1.8843487039273626e-05,
+      "loss": 0.61,
+      "step": 2320
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3166800647166592,
+      "learning_rate": 1.884231377762547e-05,
+      "loss": 0.6964,
+      "step": 2321
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3207310571794146,
+      "learning_rate": 1.884113995771643e-05,
+      "loss": 0.6904,
+      "step": 2322
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2725567726705949,
+      "learning_rate": 1.883996557962061e-05,
+      "loss": 0.6333,
+      "step": 2323
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2219432665396297,
+      "learning_rate": 1.8838790643412152e-05,
+      "loss": 0.6419,
+      "step": 2324
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.351336112935263,
+      "learning_rate": 1.883761514916524e-05,
+      "loss": 0.6341,
+      "step": 2325
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3525928252514245,
+      "learning_rate": 1.8836439096954086e-05,
+      "loss": 0.6883,
+      "step": 2326
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2945120552771745,
+      "learning_rate": 1.8835262486852944e-05,
+      "loss": 0.6079,
+      "step": 2327
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2097145550165276,
+      "learning_rate": 1.8834085318936096e-05,
+      "loss": 0.5955,
+      "step": 2328
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2838586442210311,
+      "learning_rate": 1.8832907593277868e-05,
+      "loss": 0.6218,
+      "step": 2329
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2159462001845165,
+      "learning_rate": 1.8831729309952605e-05,
+      "loss": 0.6233,
+      "step": 2330
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3050101789172515,
+      "learning_rate": 1.883055046903471e-05,
+      "loss": 0.6007,
+      "step": 2331
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.257966500089446,
+      "learning_rate": 1.8829371070598604e-05,
+      "loss": 0.5754,
+      "step": 2332
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2911298415637713,
+      "learning_rate": 1.8828191114718747e-05,
+      "loss": 0.6959,
+      "step": 2333
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2363132277038633,
+      "learning_rate": 1.8827010601469634e-05,
+      "loss": 0.5912,
+      "step": 2334
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3521765488211308,
+      "learning_rate": 1.882582953092581e-05,
+      "loss": 0.6586,
+      "step": 2335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.117424466270734,
+      "learning_rate": 1.8824647903161824e-05,
+      "loss": 0.5572,
+      "step": 2336
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2016136783569273,
+      "learning_rate": 1.882346571825229e-05,
+      "loss": 0.6283,
+      "step": 2337
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2291891559332284,
+      "learning_rate": 1.882228297627184e-05,
+      "loss": 0.6335,
+      "step": 2338
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3391902653423522,
+      "learning_rate": 1.882109967729515e-05,
+      "loss": 0.6329,
+      "step": 2339
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2842046118808663,
+      "learning_rate": 1.881991582139693e-05,
+      "loss": 0.5719,
+      "step": 2340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.338455192454326,
+      "learning_rate": 1.8818731408651914e-05,
+      "loss": 0.69,
+      "step": 2341
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3343924349286822,
+      "learning_rate": 1.8817546439134883e-05,
+      "loss": 0.6427,
+      "step": 2342
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3439261742365523,
+      "learning_rate": 1.881636091292066e-05,
+      "loss": 0.6574,
+      "step": 2343
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.379812704520425,
+      "learning_rate": 1.8815174830084084e-05,
+      "loss": 0.6522,
+      "step": 2344
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.329980384495578,
+      "learning_rate": 1.881398819070004e-05,
+      "loss": 0.6854,
+      "step": 2345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.31596865299829,
+      "learning_rate": 1.8812800994843446e-05,
+      "loss": 0.6494,
+      "step": 2346
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3828628013643132,
+      "learning_rate": 1.8811613242589257e-05,
+      "loss": 0.6334,
+      "step": 2347
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3122544285966038,
+      "learning_rate": 1.8810424934012464e-05,
+      "loss": 0.6726,
+      "step": 2348
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4490422143888047,
+      "learning_rate": 1.880923606918809e-05,
+      "loss": 0.6331,
+      "step": 2349
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2305817567309718,
+      "learning_rate": 1.8808046648191193e-05,
+      "loss": 0.6084,
+      "step": 2350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.406349771986899,
+      "learning_rate": 1.8806856671096866e-05,
+      "loss": 0.6066,
+      "step": 2351
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4005646537639251,
+      "learning_rate": 1.880566613798024e-05,
+      "loss": 0.7023,
+      "step": 2352
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.153702842848037,
+      "learning_rate": 1.880447504891648e-05,
+      "loss": 0.5712,
+      "step": 2353
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1825560499621932,
+      "learning_rate": 1.8803283403980784e-05,
+      "loss": 0.6223,
+      "step": 2354
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3774579793263662,
+      "learning_rate": 1.8802091203248388e-05,
+      "loss": 0.6601,
+      "step": 2355
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1526030167983166,
+      "learning_rate": 1.880089844679456e-05,
+      "loss": 0.5768,
+      "step": 2356
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.30984746743837,
+      "learning_rate": 1.879970513469461e-05,
+      "loss": 0.6519,
+      "step": 2357
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3555014218836876,
+      "learning_rate": 1.8798511267023874e-05,
+      "loss": 0.6513,
+      "step": 2358
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2281907566317205,
+      "learning_rate": 1.8797316843857723e-05,
+      "loss": 0.566,
+      "step": 2359
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.382401701690258,
+      "learning_rate": 1.8796121865271578e-05,
+      "loss": 0.6631,
+      "step": 2360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1713213820878148,
+      "learning_rate": 1.8794926331340874e-05,
+      "loss": 0.5769,
+      "step": 2361
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3037626906277462,
+      "learning_rate": 1.8793730242141093e-05,
+      "loss": 0.6527,
+      "step": 2362
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2611793333432209,
+      "learning_rate": 1.8792533597747756e-05,
+      "loss": 0.6272,
+      "step": 2363
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.297287840797995,
+      "learning_rate": 1.879133639823641e-05,
+      "loss": 0.5832,
+      "step": 2364
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3728048835749782,
+      "learning_rate": 1.8790138643682633e-05,
+      "loss": 0.6323,
+      "step": 2365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1537831772221736,
+      "learning_rate": 1.878894033416206e-05,
+      "loss": 0.581,
+      "step": 2366
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3256932604245857,
+      "learning_rate": 1.8787741469750332e-05,
+      "loss": 0.6431,
+      "step": 2367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2715987990498159,
+      "learning_rate": 1.8786542050523152e-05,
+      "loss": 0.6142,
+      "step": 2368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3999320201036762,
+      "learning_rate": 1.8785342076556236e-05,
+      "loss": 0.6424,
+      "step": 2369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3710288877441834,
+      "learning_rate": 1.878414154792535e-05,
+      "loss": 0.7017,
+      "step": 2370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.368623731180025,
+      "learning_rate": 1.878294046470629e-05,
+      "loss": 0.6635,
+      "step": 2371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3156077965995685,
+      "learning_rate": 1.878173882697488e-05,
+      "loss": 0.5893,
+      "step": 2372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2967169332118074,
+      "learning_rate": 1.8780536634806995e-05,
+      "loss": 0.626,
+      "step": 2373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1316281671862123,
+      "learning_rate": 1.8779333888278524e-05,
+      "loss": 0.5864,
+      "step": 2374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2839720119325606,
+      "learning_rate": 1.8778130587465414e-05,
+      "loss": 0.6651,
+      "step": 2375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1422180129957835,
+      "learning_rate": 1.877692673244363e-05,
+      "loss": 0.6014,
+      "step": 2376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.275558343815793,
+      "learning_rate": 1.877572232328918e-05,
+      "loss": 0.5993,
+      "step": 2377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1297881697546033,
+      "learning_rate": 1.8774517360078098e-05,
+      "loss": 0.5735,
+      "step": 2378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2213090781190319,
+      "learning_rate": 1.877331184288647e-05,
+      "loss": 0.5763,
+      "step": 2379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2156443518346145,
+      "learning_rate": 1.8772105771790397e-05,
+      "loss": 0.5928,
+      "step": 2380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2991646466943985,
+      "learning_rate": 1.877089914686603e-05,
+      "loss": 0.6339,
+      "step": 2381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2946729694115477,
+      "learning_rate": 1.8769691968189548e-05,
+      "loss": 0.638,
+      "step": 2382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2586145627783092,
+      "learning_rate": 1.876848423583717e-05,
+      "loss": 0.5576,
+      "step": 2383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2176354031006298,
+      "learning_rate": 1.8767275949885136e-05,
+      "loss": 0.6671,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2821997402201015,
+      "learning_rate": 1.876606711040974e-05,
+      "loss": 0.6781,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.226713912268256,
+      "learning_rate": 1.8764857717487304e-05,
+      "loss": 0.607,
+      "step": 2386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2588977751979855,
+      "learning_rate": 1.8763647771194177e-05,
+      "loss": 0.6237,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2773613884784183,
+      "learning_rate": 1.8762437271606752e-05,
+      "loss": 0.6581,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2282736438359942,
+      "learning_rate": 1.8761226218801455e-05,
+      "loss": 0.5936,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.267149442766819,
+      "learning_rate": 1.8760014612854746e-05,
+      "loss": 0.6818,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.32343568407675,
+      "learning_rate": 1.8758802453843112e-05,
+      "loss": 0.6849,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1820451028419297,
+      "learning_rate": 1.8757589741843095e-05,
+      "loss": 0.6348,
+      "step": 2392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.258630758886283,
+      "learning_rate": 1.8756376476931252e-05,
+      "loss": 0.6219,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2469574138586004,
+      "learning_rate": 1.8755162659184186e-05,
+      "loss": 0.6137,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2272257864127885,
+      "learning_rate": 1.8753948288678533e-05,
+      "loss": 0.6553,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2942705314323744,
+      "learning_rate": 1.8752733365490957e-05,
+      "loss": 0.631,
+      "step": 2396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2651904737741642,
+      "learning_rate": 1.875151788969817e-05,
+      "loss": 0.6386,
+      "step": 2397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2810628335597505,
+      "learning_rate": 1.8750301861376903e-05,
+      "loss": 0.699,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2614042758252852,
+      "learning_rate": 1.8749085280603935e-05,
+      "loss": 0.6301,
+      "step": 2399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2883318593278739,
+      "learning_rate": 1.8747868147456072e-05,
+      "loss": 0.63,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2328140916318495,
+      "learning_rate": 1.874665046201016e-05,
+      "loss": 0.632,
+      "step": 2401
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3114522657589867,
+      "learning_rate": 1.8745432224343082e-05,
+      "loss": 0.6699,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2232823338042742,
+      "learning_rate": 1.874421343453174e-05,
+      "loss": 0.5913,
+      "step": 2403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3775435676538899,
+      "learning_rate": 1.8742994092653096e-05,
+      "loss": 0.6376,
+      "step": 2404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1533099245900462,
+      "learning_rate": 1.874177419878412e-05,
+      "loss": 0.5181,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2966567166019398,
+      "learning_rate": 1.8740553753001844e-05,
+      "loss": 0.6079,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.179352415886464,
+      "learning_rate": 1.873933275538331e-05,
+      "loss": 0.5939,
+      "step": 2407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2199868015450879,
+      "learning_rate": 1.8738111206005615e-05,
+      "loss": 0.5642,
+      "step": 2408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2386872982627628,
+      "learning_rate": 1.8736889104945874e-05,
+      "loss": 0.6102,
+      "step": 2409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3689206718048275,
+      "learning_rate": 1.8735666452281246e-05,
+      "loss": 0.677,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.320869689384897,
+      "learning_rate": 1.8734443248088926e-05,
+      "loss": 0.6654,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3809177758629052,
+      "learning_rate": 1.873321949244614e-05,
+      "loss": 0.6641,
+      "step": 2412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3519685168466142,
+      "learning_rate": 1.8731995185430148e-05,
+      "loss": 0.5993,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2307249256673876,
+      "learning_rate": 1.8730770327118254e-05,
+      "loss": 0.6285,
+      "step": 2414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2267889797780407,
+      "learning_rate": 1.8729544917587778e-05,
+      "loss": 0.6415,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2773785600964451,
+      "learning_rate": 1.8728318956916096e-05,
+      "loss": 0.5713,
+      "step": 2416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2343488762903918,
+      "learning_rate": 1.8727092445180605e-05,
+      "loss": 0.5605,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4341248404244182,
+      "learning_rate": 1.8725865382458744e-05,
+      "loss": 0.6518,
+      "step": 2418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3182418205691848,
+      "learning_rate": 1.8724637768827982e-05,
+      "loss": 0.632,
+      "step": 2419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1458106067600007,
+      "learning_rate": 1.8723409604365823e-05,
+      "loss": 0.5761,
+      "step": 2420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.330170525969723,
+      "learning_rate": 1.872218088914981e-05,
+      "loss": 0.6906,
+      "step": 2421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2021378269548495,
+      "learning_rate": 1.8720951623257514e-05,
+      "loss": 0.6223,
+      "step": 2422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1857673150989387,
+      "learning_rate": 1.871972180676655e-05,
+      "loss": 0.5929,
+      "step": 2423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2485561137851693,
+      "learning_rate": 1.8718491439754562e-05,
+      "loss": 0.6092,
+      "step": 2424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.275888342955678,
+      "learning_rate": 1.8717260522299226e-05,
+      "loss": 0.6699,
+      "step": 2425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2705311754733652,
+      "learning_rate": 1.8716029054478257e-05,
+      "loss": 0.6836,
+      "step": 2426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2274701593038857,
+      "learning_rate": 1.8714797036369407e-05,
+      "loss": 0.6395,
+      "step": 2427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1634775522588696,
+      "learning_rate": 1.8713564468050455e-05,
+      "loss": 0.5808,
+      "step": 2428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.237808955764608,
+      "learning_rate": 1.8712331349599227e-05,
+      "loss": 0.6695,
+      "step": 2429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2227718798283174,
+      "learning_rate": 1.8711097681093568e-05,
+      "loss": 0.6011,
+      "step": 2430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1968562809448942,
+      "learning_rate": 1.8709863462611368e-05,
+      "loss": 0.6742,
+      "step": 2431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4422217235864745,
+      "learning_rate": 1.8708628694230554e-05,
+      "loss": 0.6482,
+      "step": 2432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2474815747996928,
+      "learning_rate": 1.8707393376029076e-05,
+      "loss": 0.6287,
+      "step": 2433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1562748983640754,
+      "learning_rate": 1.8706157508084934e-05,
+      "loss": 0.6359,
+      "step": 2434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2451794178688973,
+      "learning_rate": 1.8704921090476148e-05,
+      "loss": 0.6593,
+      "step": 2435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.232030017298225,
+      "learning_rate": 1.870368412328078e-05,
+      "loss": 0.6008,
+      "step": 2436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.139302623749709,
+      "learning_rate": 1.870244660657693e-05,
+      "loss": 0.5903,
+      "step": 2437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2632565412776657,
+      "learning_rate": 1.870120854044273e-05,
+      "loss": 0.6103,
+      "step": 2438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3128410532024872,
+      "learning_rate": 1.869996992495634e-05,
+      "loss": 0.6331,
+      "step": 2439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3096785418841905,
+      "learning_rate": 1.8698730760195963e-05,
+      "loss": 0.6391,
+      "step": 2440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2535515874861367,
+      "learning_rate": 1.869749104623983e-05,
+      "loss": 0.6135,
+      "step": 2441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2570058000926945,
+      "learning_rate": 1.8696250783166218e-05,
+      "loss": 0.6095,
+      "step": 2442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2589870678916961,
+      "learning_rate": 1.869500997105343e-05,
+      "loss": 0.6358,
+      "step": 2443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1986070992874267,
+      "learning_rate": 1.8693768609979796e-05,
+      "loss": 0.5585,
+      "step": 2444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1968443286441857,
+      "learning_rate": 1.86925267000237e-05,
+      "loss": 0.6206,
+      "step": 2445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.22023404073097,
+      "learning_rate": 1.8691284241263547e-05,
+      "loss": 0.6005,
+      "step": 2446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2934050500950776,
+      "learning_rate": 1.8690041233777774e-05,
+      "loss": 0.662,
+      "step": 2447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2180060047571653,
+      "learning_rate": 1.8688797677644865e-05,
+      "loss": 0.5908,
+      "step": 2448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2152569426915234,
+      "learning_rate": 1.868755357294333e-05,
+      "loss": 0.662,
+      "step": 2449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4536301749942062,
+      "learning_rate": 1.8686308919751718e-05,
+      "loss": 0.6754,
+      "step": 2450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2810053708144795,
+      "learning_rate": 1.8685063718148608e-05,
+      "loss": 0.5837,
+      "step": 2451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3048902799440902,
+      "learning_rate": 1.8683817968212613e-05,
+      "loss": 0.713,
+      "step": 2452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.275016180571564,
+      "learning_rate": 1.868257167002239e-05,
+      "loss": 0.6437,
+      "step": 2453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1846074965664053,
+      "learning_rate": 1.8681324823656623e-05,
+      "loss": 0.5569,
+      "step": 2454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2770317699898526,
+      "learning_rate": 1.8680077429194025e-05,
+      "loss": 0.6058,
+      "step": 2455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3043310626480014,
+      "learning_rate": 1.8678829486713362e-05,
+      "loss": 0.7058,
+      "step": 2456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.462291151605452,
+      "learning_rate": 1.867758099629341e-05,
+      "loss": 0.6179,
+      "step": 2457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2707998656988748,
+      "learning_rate": 1.8676331958013004e-05,
+      "loss": 0.6079,
+      "step": 2458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2663154602452986,
+      "learning_rate": 1.8675082371950996e-05,
+      "loss": 0.6044,
+      "step": 2459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1410610462878796,
+      "learning_rate": 1.867383223818628e-05,
+      "loss": 0.5609,
+      "step": 2460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3634408677011673,
+      "learning_rate": 1.8672581556797785e-05,
+      "loss": 0.6513,
+      "step": 2461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2271815396157595,
+      "learning_rate": 1.867133032786447e-05,
+      "loss": 0.6535,
+      "step": 2462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2678103476701896,
+      "learning_rate": 1.8670078551465336e-05,
+      "loss": 0.6123,
+      "step": 2463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2029035290083356,
+      "learning_rate": 1.8668826227679408e-05,
+      "loss": 0.5776,
+      "step": 2464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2136313714966078,
+      "learning_rate": 1.8667573356585756e-05,
+      "loss": 0.6221,
+      "step": 2465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3528623559497148,
+      "learning_rate": 1.866631993826348e-05,
+      "loss": 0.6203,
+      "step": 2466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.220765818702567,
+      "learning_rate": 1.8665065972791715e-05,
+      "loss": 0.6093,
+      "step": 2467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3541523370229156,
+      "learning_rate": 1.8663811460249625e-05,
+      "loss": 0.6144,
+      "step": 2468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2202700403619804,
+      "learning_rate": 1.8662556400716423e-05,
+      "loss": 0.6005,
+      "step": 2469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.277062901410712,
+      "learning_rate": 1.866130079427134e-05,
+      "loss": 0.5709,
+      "step": 2470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2106963378968074,
+      "learning_rate": 1.866004464099365e-05,
+      "loss": 0.6306,
+      "step": 2471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.270141361058738,
+      "learning_rate": 1.8658787940962666e-05,
+      "loss": 0.6145,
+      "step": 2472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2725296060531797,
+      "learning_rate": 1.8657530694257722e-05,
+      "loss": 0.6286,
+      "step": 2473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1505708666597803,
+      "learning_rate": 1.8656272900958202e-05,
+      "loss": 0.5983,
+      "step": 2474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.211748503781468,
+      "learning_rate": 1.865501456114351e-05,
+      "loss": 0.6194,
+      "step": 2475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.246196395807835,
+      "learning_rate": 1.8653755674893095e-05,
+      "loss": 0.6014,
+      "step": 2476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2684632489357968,
+      "learning_rate": 1.865249624228644e-05,
+      "loss": 0.6523,
+      "step": 2477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2319233365110043,
+      "learning_rate": 1.8651236263403052e-05,
+      "loss": 0.6185,
+      "step": 2478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2574923090007528,
+      "learning_rate": 1.8649975738322487e-05,
+      "loss": 0.5813,
+      "step": 2479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2989334866804474,
+      "learning_rate": 1.8648714667124323e-05,
+      "loss": 0.6515,
+      "step": 2480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1623842571434297,
+      "learning_rate": 1.8647453049888186e-05,
+      "loss": 0.5784,
+      "step": 2481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2289409537971798,
+      "learning_rate": 1.8646190886693717e-05,
+      "loss": 0.584,
+      "step": 2482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.148413158340434,
+      "learning_rate": 1.8644928177620612e-05,
+      "loss": 0.571,
+      "step": 2483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2768933730033598,
+      "learning_rate": 1.864366492274859e-05,
+      "loss": 0.6365,
+      "step": 2484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3055138626385328,
+      "learning_rate": 1.86424011221574e-05,
+      "loss": 0.5909,
+      "step": 2485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2036296232092925,
+      "learning_rate": 1.8641136775926844e-05,
+      "loss": 0.6286,
+      "step": 2486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1856331958799164,
+      "learning_rate": 1.863987188413674e-05,
+      "loss": 0.5756,
+      "step": 2487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.134590962767589,
+      "learning_rate": 1.8638606446866947e-05,
+      "loss": 0.4757,
+      "step": 2488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2513143781626266,
+      "learning_rate": 1.8637340464197357e-05,
+      "loss": 0.6306,
+      "step": 2489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1773843577266503,
+      "learning_rate": 1.8636073936207907e-05,
+      "loss": 0.5852,
+      "step": 2490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1947211005544542,
+      "learning_rate": 1.863480686297855e-05,
+      "loss": 0.66,
+      "step": 2491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2703404591241307,
+      "learning_rate": 1.8633539244589285e-05,
+      "loss": 0.5839,
+      "step": 2492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.310701864205226,
+      "learning_rate": 1.8632271081120144e-05,
+      "loss": 0.5919,
+      "step": 2493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2219158526759577,
+      "learning_rate": 1.8631002372651194e-05,
+      "loss": 0.6121,
+      "step": 2494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.25539051279771,
+      "learning_rate": 1.8629733119262536e-05,
+      "loss": 0.61,
+      "step": 2495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1989805500657778,
+      "learning_rate": 1.86284633210343e-05,
+      "loss": 0.5397,
+      "step": 2496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.220097066347891,
+      "learning_rate": 1.862719297804666e-05,
+      "loss": 0.6262,
+      "step": 2497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3601523621145881,
+      "learning_rate": 1.8625922090379812e-05,
+      "loss": 0.6197,
+      "step": 2498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.261808173727957,
+      "learning_rate": 1.8624650658114003e-05,
+      "loss": 0.6342,
+      "step": 2499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.097432232461151,
+      "learning_rate": 1.86233786813295e-05,
+      "loss": 0.5538,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4307810452386007,
+      "learning_rate": 1.862210616010661e-05,
+      "loss": 0.7265,
+      "step": 2501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2982976532874413,
+      "learning_rate": 1.862083309452567e-05,
+      "loss": 0.6277,
+      "step": 2502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2739126804975092,
+      "learning_rate": 1.8619559484667063e-05,
+      "loss": 0.6756,
+      "step": 2503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2373585918821108,
+      "learning_rate": 1.8618285330611193e-05,
+      "loss": 0.6096,
+      "step": 2504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3507258741987285,
+      "learning_rate": 1.8617010632438508e-05,
+      "loss": 0.6555,
+      "step": 2505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.116901761307012,
+      "learning_rate": 1.8615735390229483e-05,
+      "loss": 0.6206,
+      "step": 2506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1878597818752807,
+      "learning_rate": 1.8614459604064634e-05,
+      "loss": 0.5414,
+      "step": 2507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2555094894512393,
+      "learning_rate": 1.86131832740245e-05,
+      "loss": 0.6433,
+      "step": 2508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2274999741284276,
+      "learning_rate": 1.861190640018967e-05,
+      "loss": 0.6259,
+      "step": 2509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2973945220528185,
+      "learning_rate": 1.8610628982640758e-05,
+      "loss": 0.6293,
+      "step": 2510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1606728787424836,
+      "learning_rate": 1.8609351021458412e-05,
+      "loss": 0.5604,
+      "step": 2511
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3960318732571626,
+      "learning_rate": 1.8608072516723317e-05,
+      "loss": 0.6043,
+      "step": 2512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3159074602432916,
+      "learning_rate": 1.8606793468516197e-05,
+      "loss": 0.625,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2911882386111526,
+      "learning_rate": 1.8605513876917793e-05,
+      "loss": 0.6286,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1841470161774408,
+      "learning_rate": 1.8604233742008903e-05,
+      "loss": 0.6274,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.158395169114741,
+      "learning_rate": 1.860295306387035e-05,
+      "loss": 0.6291,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1611165911004209,
+      "learning_rate": 1.8601671842582977e-05,
+      "loss": 0.5685,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2891511828810491,
+      "learning_rate": 1.8600390078227682e-05,
+      "loss": 0.6666,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2628273832165804,
+      "learning_rate": 1.8599107770885392e-05,
+      "loss": 0.5996,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2303213848697214,
+      "learning_rate": 1.8597824920637063e-05,
+      "loss": 0.6127,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2173923364147705,
+      "learning_rate": 1.8596541527563686e-05,
+      "loss": 0.5605,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1828597912222343,
+      "learning_rate": 1.859525759174629e-05,
+      "loss": 0.6037,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3533218927714155,
+      "learning_rate": 1.8593973113265938e-05,
+      "loss": 0.5849,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1939991735874045,
+      "learning_rate": 1.8592688092203727e-05,
+      "loss": 0.5959,
+      "step": 2524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3305723641638398,
+      "learning_rate": 1.8591402528640782e-05,
+      "loss": 0.6473,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2634680464205919,
+      "learning_rate": 1.8590116422658267e-05,
+      "loss": 0.5907,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2424357903850305,
+      "learning_rate": 1.8588829774337388e-05,
+      "loss": 0.6681,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.290836708494584,
+      "learning_rate": 1.858754258375937e-05,
+      "loss": 0.6205,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1172271934873748,
+      "learning_rate": 1.858625485100548e-05,
+      "loss": 0.5876,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3565895657496478,
+      "learning_rate": 1.8584966576157028e-05,
+      "loss": 0.5838,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1605733513420058,
+      "learning_rate": 1.8583677759295337e-05,
+      "loss": 0.5603,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1943168732189107,
+      "learning_rate": 1.8582388400501786e-05,
+      "loss": 0.5914,
+      "step": 2532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.340174775895629,
+      "learning_rate": 1.8581098499857776e-05,
+      "loss": 0.636,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.195819871718148,
+      "learning_rate": 1.8579808057444744e-05,
+      "loss": 0.6409,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3152215533853757,
+      "learning_rate": 1.8578517073344168e-05,
+      "loss": 0.5781,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1787502784769128,
+      "learning_rate": 1.8577225547637545e-05,
+      "loss": 0.6261,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2210191485460118,
+      "learning_rate": 1.857593348040642e-05,
+      "loss": 0.6121,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2325777901325448,
+      "learning_rate": 1.857464087173237e-05,
+      "loss": 0.6224,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2258120628359654,
+      "learning_rate": 1.8573347721697007e-05,
+      "loss": 0.5919,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2111445588315997,
+      "learning_rate": 1.8572054030381963e-05,
+      "loss": 0.5985,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2401532003916247,
+      "learning_rate": 1.8570759797868925e-05,
+      "loss": 0.6971,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1244443474885986,
+      "learning_rate": 1.8569465024239603e-05,
+      "loss": 0.5731,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1371780463882148,
+      "learning_rate": 1.8568169709575737e-05,
+      "loss": 0.6314,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2121776506260318,
+      "learning_rate": 1.856687385395911e-05,
+      "loss": 0.6172,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.359451423338358,
+      "learning_rate": 1.8565577457471545e-05,
+      "loss": 0.6485,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2032117936175215,
+      "learning_rate": 1.8564280520194876e-05,
+      "loss": 0.5711,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.418571743290778,
+      "learning_rate": 1.8562983042210998e-05,
+      "loss": 0.6426,
+      "step": 2547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3211135691324105,
+      "learning_rate": 1.8561685023601815e-05,
+      "loss": 0.5697,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3007259414415762,
+      "learning_rate": 1.8560386464449288e-05,
+      "loss": 0.6598,
+      "step": 2549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1542516915761523,
+      "learning_rate": 1.8559087364835397e-05,
+      "loss": 0.6174,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3336154321428684,
+      "learning_rate": 1.855778772484216e-05,
+      "loss": 0.6085,
+      "step": 2551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.269227023719383,
+      "learning_rate": 1.8556487544551633e-05,
+      "loss": 0.5854,
+      "step": 2552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3994029883873589,
+      "learning_rate": 1.8555186824045903e-05,
+      "loss": 0.6521,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3745360458687568,
+      "learning_rate": 1.8553885563407086e-05,
+      "loss": 0.6457,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1461255423433991,
+      "learning_rate": 1.8552583762717346e-05,
+      "loss": 0.5572,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2501790395308439,
+      "learning_rate": 1.8551281422058863e-05,
+      "loss": 0.6272,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.299685959297986,
+      "learning_rate": 1.8549978541513865e-05,
+      "loss": 0.6544,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1721466766790667,
+      "learning_rate": 1.8548675121164613e-05,
+      "loss": 0.625,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2346274685102632,
+      "learning_rate": 1.8547371161093397e-05,
+      "loss": 0.6582,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.145131237055101,
+      "learning_rate": 1.8546066661382535e-05,
+      "loss": 0.5911,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1776541560800873,
+      "learning_rate": 1.8544761622114396e-05,
+      "loss": 0.6055,
+      "step": 2561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3578178870968904,
+      "learning_rate": 1.854345604337137e-05,
+      "loss": 0.6334,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1676885976947013,
+      "learning_rate": 1.8542149925235885e-05,
+      "loss": 0.5701,
+      "step": 2563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2746782401745258,
+      "learning_rate": 1.8540843267790406e-05,
+      "loss": 0.6239,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1632292152382684,
+      "learning_rate": 1.8539536071117424e-05,
+      "loss": 0.5816,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1965253573232237,
+      "learning_rate": 1.853822833529947e-05,
+      "loss": 0.5795,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1052932869597845,
+      "learning_rate": 1.853692006041911e-05,
+      "loss": 0.589,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2843409679943978,
+      "learning_rate": 1.8535611246558947e-05,
+      "loss": 0.6422,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2733504844845251,
+      "learning_rate": 1.8534301893801606e-05,
+      "loss": 0.6569,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2016262776645703,
+      "learning_rate": 1.8532992002229755e-05,
+      "loss": 0.6028,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.420376753324155,
+      "learning_rate": 1.853168157192609e-05,
+      "loss": 0.6932,
+      "step": 2571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1569254165454455,
+      "learning_rate": 1.8530370602973356e-05,
+      "loss": 0.5704,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3356085924525085,
+      "learning_rate": 1.852905909545431e-05,
+      "loss": 0.6361,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1938464580132024,
+      "learning_rate": 1.8527747049451763e-05,
+      "loss": 0.6223,
+      "step": 2574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2820819386766096,
+      "learning_rate": 1.8526434465048544e-05,
+      "loss": 0.6493,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2741980592508404,
+      "learning_rate": 1.852512134232753e-05,
+      "loss": 0.6682,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1854992123218688,
+      "learning_rate": 1.8523807681371622e-05,
+      "loss": 0.6324,
+      "step": 2577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.340573081765185,
+      "learning_rate": 1.8522493482263753e-05,
+      "loss": 0.6448,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.214321993438132,
+      "learning_rate": 1.8521178745086906e-05,
+      "loss": 0.6302,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1534556051903955,
+      "learning_rate": 1.8519863469924078e-05,
+      "loss": 0.5599,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2511316422190768,
+      "learning_rate": 1.8518547656858314e-05,
+      "loss": 0.6031,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.309353963459886,
+      "learning_rate": 1.8517231305972686e-05,
+      "loss": 0.6538,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1116357789807634,
+      "learning_rate": 1.85159144173503e-05,
+      "loss": 0.5839,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3377808463173022,
+      "learning_rate": 1.8514596991074305e-05,
+      "loss": 0.6171,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2714148530250609,
+      "learning_rate": 1.8513279027227868e-05,
+      "loss": 0.6274,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1845584878229116,
+      "learning_rate": 1.8511960525894207e-05,
+      "loss": 0.5645,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2666671135968541,
+      "learning_rate": 1.851064148715656e-05,
+      "loss": 0.6043,
+      "step": 2587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1955396149731525,
+      "learning_rate": 1.8509321911098206e-05,
+      "loss": 0.5603,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3599628186051949,
+      "learning_rate": 1.850800179780246e-05,
+      "loss": 0.664,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1808500309106609,
+      "learning_rate": 1.8506681147352662e-05,
+      "loss": 0.5671,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1913907409659108,
+      "learning_rate": 1.8505359959832198e-05,
+      "loss": 0.6267,
+      "step": 2591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2191746289889973,
+      "learning_rate": 1.8504038235324474e-05,
+      "loss": 0.644,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1284596969039795,
+      "learning_rate": 1.8502715973912942e-05,
+      "loss": 0.5789,
+      "step": 2593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.272040367154707,
+      "learning_rate": 1.8501393175681084e-05,
+      "loss": 0.6216,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2028242453648108,
+      "learning_rate": 1.8500069840712412e-05,
+      "loss": 0.5937,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1447960348743038,
+      "learning_rate": 1.849874596909048e-05,
+      "loss": 0.5663,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.221488224093012,
+      "learning_rate": 1.849742156089886e-05,
+      "loss": 0.6107,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3577475616764714,
+      "learning_rate": 1.8496096616221182e-05,
+      "loss": 0.65,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2733652761373986,
+      "learning_rate": 1.8494771135141086e-05,
+      "loss": 0.5847,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3544261341449815,
+      "learning_rate": 1.8493445117742267e-05,
+      "loss": 0.7011,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2235235970750573,
+      "learning_rate": 1.8492118564108433e-05,
+      "loss": 0.6319,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2630315039534614,
+      "learning_rate": 1.849079147432334e-05,
+      "loss": 0.6244,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2065669280837659,
+      "learning_rate": 1.8489463848470777e-05,
+      "loss": 0.5747,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2684779096220387,
+      "learning_rate": 1.848813568663456e-05,
+      "loss": 0.6744,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2708911830471694,
+      "learning_rate": 1.8486806988898545e-05,
+      "loss": 0.6292,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.387470954728476,
+      "learning_rate": 1.848547775534662e-05,
+      "loss": 0.6168,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.21128894259088,
+      "learning_rate": 1.8484147986062702e-05,
+      "loss": 0.5725,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.234802606614738,
+      "learning_rate": 1.8482817681130752e-05,
+      "loss": 0.6094,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2190169017705899,
+      "learning_rate": 1.8481486840634753e-05,
+      "loss": 0.6704,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.258088123143036,
+      "learning_rate": 1.848015546465873e-05,
+      "loss": 0.6575,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.215394756802738,
+      "learning_rate": 1.8478823553286745e-05,
+      "loss": 0.5818,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2669526189391536,
+      "learning_rate": 1.847749110660288e-05,
+      "loss": 0.6307,
+      "step": 2612
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.091879143797114,
+      "learning_rate": 1.847615812469127e-05,
+      "loss": 0.571,
+      "step": 2613
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1244845269013442,
+      "learning_rate": 1.847482460763606e-05,
+      "loss": 0.5366,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1921238973507857,
+      "learning_rate": 1.847349055552145e-05,
+      "loss": 0.5979,
+      "step": 2615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2410866039559778,
+      "learning_rate": 1.8472155968431664e-05,
+      "loss": 0.6689,
+      "step": 2616
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.356489912690882,
+      "learning_rate": 1.847082084645096e-05,
+      "loss": 0.7003,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2204231124157234,
+      "learning_rate": 1.8469485189663635e-05,
+      "loss": 0.6102,
+      "step": 2618
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2619691015774441,
+      "learning_rate": 1.846814899815401e-05,
+      "loss": 0.6076,
+      "step": 2619
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1410021746194294,
+      "learning_rate": 1.846681227200645e-05,
+      "loss": 0.6001,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.271208514921696,
+      "learning_rate": 1.846547501130535e-05,
+      "loss": 0.6767,
+      "step": 2621
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0982250219744067,
+      "learning_rate": 1.8464137216135133e-05,
+      "loss": 0.5341,
+      "step": 2622
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.265605549603989,
+      "learning_rate": 1.846279888658027e-05,
+      "loss": 0.6375,
+      "step": 2623
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2173059173294758,
+      "learning_rate": 1.8461460022725247e-05,
+      "loss": 0.6639,
+      "step": 2624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.282403101392959,
+      "learning_rate": 1.8460120624654595e-05,
+      "loss": 0.5959,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2232820901783792,
+      "learning_rate": 1.8458780692452882e-05,
+      "loss": 0.6204,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2671190556152214,
+      "learning_rate": 1.8457440226204702e-05,
+      "loss": 0.6453,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2586802457027126,
+      "learning_rate": 1.8456099225994687e-05,
+      "loss": 0.6056,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2801579263984155,
+      "learning_rate": 1.84547576919075e-05,
+      "loss": 0.6207,
+      "step": 2629
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2125485931569533,
+      "learning_rate": 1.845341562402784e-05,
+      "loss": 0.6263,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2536828623449645,
+      "learning_rate": 1.8452073022440436e-05,
+      "loss": 0.6718,
+      "step": 2631
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.29967642021547,
+      "learning_rate": 1.8450729887230053e-05,
+      "loss": 0.6593,
+      "step": 2632
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2974322856660658,
+      "learning_rate": 1.8449386218481495e-05,
+      "loss": 0.6644,
+      "step": 2633
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3146686348945504,
+      "learning_rate": 1.8448042016279592e-05,
+      "loss": 0.6526,
+      "step": 2634
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.327386987620711,
+      "learning_rate": 1.844669728070921e-05,
+      "loss": 0.6544,
+      "step": 2635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.12063802036055,
+      "learning_rate": 1.844535201185525e-05,
+      "loss": 0.5968,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3718246294302483,
+      "learning_rate": 1.8444006209802646e-05,
+      "loss": 0.6414,
+      "step": 2637
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0747232258264159,
+      "learning_rate": 1.8442659874636365e-05,
+      "loss": 0.5564,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1955376705933867,
+      "learning_rate": 1.8441313006441402e-05,
+      "loss": 0.5549,
+      "step": 2639
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2364534191022152,
+      "learning_rate": 1.8439965605302803e-05,
+      "loss": 0.5558,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.229463342111611,
+      "learning_rate": 1.843861767130563e-05,
+      "loss": 0.6207,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.301853276592066,
+      "learning_rate": 1.8437269204534983e-05,
+      "loss": 0.6566,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0536674370328925,
+      "learning_rate": 1.8435920205076003e-05,
+      "loss": 0.5349,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2264069379575435,
+      "learning_rate": 1.843457067301386e-05,
+      "loss": 0.6085,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3700773813850702,
+      "learning_rate": 1.8433220608433747e-05,
+      "loss": 0.6298,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3546180804108368,
+      "learning_rate": 1.8431870011420908e-05,
+      "loss": 0.6848,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2270667624960336,
+      "learning_rate": 1.8430518882060618e-05,
+      "loss": 0.6283,
+      "step": 2647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2408093178608892,
+      "learning_rate": 1.8429167220438172e-05,
+      "loss": 0.6122,
+      "step": 2648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.224188676086163,
+      "learning_rate": 1.842781502663891e-05,
+      "loss": 0.6537,
+      "step": 2649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2322450925141506,
+      "learning_rate": 1.8426462300748206e-05,
+      "loss": 0.6884,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2752650901621032,
+      "learning_rate": 1.842510904285146e-05,
+      "loss": 0.6021,
+      "step": 2651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.227564942687233,
+      "learning_rate": 1.8423755253034113e-05,
+      "loss": 0.6353,
+      "step": 2652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2917440093890669,
+      "learning_rate": 1.8422400931381635e-05,
+      "loss": 0.6622,
+      "step": 2653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3524081721131431,
+      "learning_rate": 1.8421046077979535e-05,
+      "loss": 0.6717,
+      "step": 2654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.243483916531045,
+      "learning_rate": 1.8419690692913346e-05,
+      "loss": 0.6144,
+      "step": 2655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.332726126730903,
+      "learning_rate": 1.8418334776268643e-05,
+      "loss": 0.6453,
+      "step": 2656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.246272202870776,
+      "learning_rate": 1.8416978328131036e-05,
+      "loss": 0.6248,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.239141604579045,
+      "learning_rate": 1.841562134858616e-05,
+      "loss": 0.6275,
+      "step": 2658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2189441917449733,
+      "learning_rate": 1.8414263837719688e-05,
+      "loss": 0.5938,
+      "step": 2659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2622618554492198,
+      "learning_rate": 1.8412905795617327e-05,
+      "loss": 0.6258,
+      "step": 2660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2496346416584134,
+      "learning_rate": 1.8411547222364824e-05,
+      "loss": 0.6185,
+      "step": 2661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2873301569937172,
+      "learning_rate": 1.8410188118047942e-05,
+      "loss": 0.6237,
+      "step": 2662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1988220055586474,
+      "learning_rate": 1.8408828482752495e-05,
+      "loss": 0.6378,
+      "step": 2663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2618972132457877,
+      "learning_rate": 1.8407468316564322e-05,
+      "loss": 0.6317,
+      "step": 2664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.140429022688034,
+      "learning_rate": 1.84061076195693e-05,
+      "loss": 0.6108,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.221287262581384,
+      "learning_rate": 1.8404746391853328e-05,
+      "loss": 0.6259,
+      "step": 2666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2791333276005448,
+      "learning_rate": 1.840338463350236e-05,
+      "loss": 0.6019,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.307468078179842,
+      "learning_rate": 1.840202234460236e-05,
+      "loss": 0.6018,
+      "step": 2668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.222592483173903,
+      "learning_rate": 1.8400659525239343e-05,
+      "loss": 0.6686,
+      "step": 2669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2743021364134406,
+      "learning_rate": 1.8399296175499346e-05,
+      "loss": 0.6374,
+      "step": 2670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0510973827060548,
+      "learning_rate": 1.8397932295468446e-05,
+      "loss": 0.5676,
+      "step": 2671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2944416979777424,
+      "learning_rate": 1.8396567885232753e-05,
+      "loss": 0.6113,
+      "step": 2672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1775259969153802,
+      "learning_rate": 1.839520294487841e-05,
+      "loss": 0.6515,
+      "step": 2673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3056745621480876,
+      "learning_rate": 1.8393837474491595e-05,
+      "loss": 0.6288,
+      "step": 2674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.243677505220659,
+      "learning_rate": 1.839247147415851e-05,
+      "loss": 0.6424,
+      "step": 2675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3332101298074706,
+      "learning_rate": 1.83911049439654e-05,
+      "loss": 0.7071,
+      "step": 2676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1380195646766325,
+      "learning_rate": 1.8389737883998545e-05,
+      "loss": 0.6372,
+      "step": 2677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2390205751424068,
+      "learning_rate": 1.838837029434425e-05,
+      "loss": 0.6337,
+      "step": 2678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4461415155684236,
+      "learning_rate": 1.8387002175088863e-05,
+      "loss": 0.6239,
+      "step": 2679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2419222662521607,
+      "learning_rate": 1.8385633526318758e-05,
+      "loss": 0.624,
+      "step": 2680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1862414868836182,
+      "learning_rate": 1.838426434812034e-05,
+      "loss": 0.6165,
+      "step": 2681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2945934586895842,
+      "learning_rate": 1.838289464058006e-05,
+      "loss": 0.6867,
+      "step": 2682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2284431375487235,
+      "learning_rate": 1.838152440378439e-05,
+      "loss": 0.6384,
+      "step": 2683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.261263502443776,
+      "learning_rate": 1.8380153637819843e-05,
+      "loss": 0.6266,
+      "step": 2684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1933221639522478,
+      "learning_rate": 1.8378782342772958e-05,
+      "loss": 0.5394,
+      "step": 2685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1678693339926849,
+      "learning_rate": 1.8377410518730318e-05,
+      "loss": 0.5672,
+      "step": 2686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2024879255104957,
+      "learning_rate": 1.8376038165778526e-05,
+      "loss": 0.6657,
+      "step": 2687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1051933565602114,
+      "learning_rate": 1.8374665284004235e-05,
+      "loss": 0.5627,
+      "step": 2688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2497066153504524,
+      "learning_rate": 1.8373291873494112e-05,
+      "loss": 0.6754,
+      "step": 2689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2209087717402778,
+      "learning_rate": 1.8371917934334877e-05,
+      "loss": 0.6259,
+      "step": 2690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2217406235583503,
+      "learning_rate": 1.8370543466613267e-05,
+      "loss": 0.5533,
+      "step": 2691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0858125683391366,
+      "learning_rate": 1.8369168470416064e-05,
+      "loss": 0.5388,
+      "step": 2692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2870784865798823,
+      "learning_rate": 1.8367792945830073e-05,
+      "loss": 0.7085,
+      "step": 2693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1734353422305082,
+      "learning_rate": 1.836641689294214e-05,
+      "loss": 0.6266,
+      "step": 2694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3411617191694807,
+      "learning_rate": 1.8365040311839148e-05,
+      "loss": 0.618,
+      "step": 2695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3621934633415456,
+      "learning_rate": 1.8363663202608e-05,
+      "loss": 0.6862,
+      "step": 2696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2500355238635081,
+      "learning_rate": 1.8362285565335645e-05,
+      "loss": 0.618,
+      "step": 2697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3374980712591507,
+      "learning_rate": 1.836090740010906e-05,
+      "loss": 0.6414,
+      "step": 2698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3028346170408762,
+      "learning_rate": 1.835952870701525e-05,
+      "loss": 0.6513,
+      "step": 2699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2482191274390388,
+      "learning_rate": 1.8358149486141265e-05,
+      "loss": 0.6213,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2262618552907405,
+      "learning_rate": 1.835676973757418e-05,
+      "loss": 0.5418,
+      "step": 2701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1754886098266861,
+      "learning_rate": 1.835538946140111e-05,
+      "loss": 0.5945,
+      "step": 2702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2011351541007982,
+      "learning_rate": 1.835400865770919e-05,
+      "loss": 0.6356,
+      "step": 2703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.258339623357618,
+      "learning_rate": 1.8352627326585606e-05,
+      "loss": 0.6318,
+      "step": 2704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2204543203473468,
+      "learning_rate": 1.8351245468117564e-05,
+      "loss": 0.6116,
+      "step": 2705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1386343424714571,
+      "learning_rate": 1.834986308239231e-05,
+      "loss": 0.686,
+      "step": 2706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1711072822995203,
+      "learning_rate": 1.8348480169497114e-05,
+      "loss": 0.573,
+      "step": 2707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1923948595092952,
+      "learning_rate": 1.8347096729519298e-05,
+      "loss": 0.5769,
+      "step": 2708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2338145951634476,
+      "learning_rate": 1.83457127625462e-05,
+      "loss": 0.5593,
+      "step": 2709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3051236731291675,
+      "learning_rate": 1.8344328268665193e-05,
+      "loss": 0.6239,
+      "step": 2710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3311047506118519,
+      "learning_rate": 1.83429432479637e-05,
+      "loss": 0.632,
+      "step": 2711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.278516178847769,
+      "learning_rate": 1.834155770052915e-05,
+      "loss": 0.5806,
+      "step": 2712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.323828580647444,
+      "learning_rate": 1.8340171626449025e-05,
+      "loss": 0.5996,
+      "step": 2713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1035350696572261,
+      "learning_rate": 1.8338785025810836e-05,
+      "loss": 0.5794,
+      "step": 2714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3069138405207896,
+      "learning_rate": 1.8337397898702127e-05,
+      "loss": 0.6397,
+      "step": 2715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2692549184689246,
+      "learning_rate": 1.833601024521047e-05,
+      "loss": 0.5903,
+      "step": 2716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2384900900729578,
+      "learning_rate": 1.8334622065423482e-05,
+      "loss": 0.6285,
+      "step": 2717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.11634362634413,
+      "learning_rate": 1.8333233359428798e-05,
+      "loss": 0.6394,
+      "step": 2718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2987121530255736,
+      "learning_rate": 1.8331844127314104e-05,
+      "loss": 0.6591,
+      "step": 2719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1814799554216615,
+      "learning_rate": 1.83304543691671e-05,
+      "loss": 0.588,
+      "step": 2720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3095366767358483,
+      "learning_rate": 1.832906408507553e-05,
+      "loss": 0.6206,
+      "step": 2721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.314702864752759,
+      "learning_rate": 1.8327673275127177e-05,
+      "loss": 0.669,
+      "step": 2722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.191442370258049,
+      "learning_rate": 1.832628193940984e-05,
+      "loss": 0.6409,
+      "step": 2723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3098278091190558,
+      "learning_rate": 1.832489007801137e-05,
+      "loss": 0.6143,
+      "step": 2724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1939614832622507,
+      "learning_rate": 1.832349769101964e-05,
+      "loss": 0.5463,
+      "step": 2725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.284174071325053,
+      "learning_rate": 1.832210477852255e-05,
+      "loss": 0.6731,
+      "step": 2726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1573902384579537,
+      "learning_rate": 1.8320711340608058e-05,
+      "loss": 0.6316,
+      "step": 2727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.210314633549534,
+      "learning_rate": 1.8319317377364124e-05,
+      "loss": 0.6184,
+      "step": 2728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.216356519247249,
+      "learning_rate": 1.8317922888878763e-05,
+      "loss": 0.6378,
+      "step": 2729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2886793549763733,
+      "learning_rate": 1.8316527875240016e-05,
+      "loss": 0.5714,
+      "step": 2730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1853238293453052,
+      "learning_rate": 1.8315132336535956e-05,
+      "loss": 0.5943,
+      "step": 2731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4009815283516742,
+      "learning_rate": 1.8313736272854692e-05,
+      "loss": 0.5872,
+      "step": 2732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3185053991757636,
+      "learning_rate": 1.8312339684284363e-05,
+      "loss": 0.7169,
+      "step": 2733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1847111931889096,
+      "learning_rate": 1.831094257091314e-05,
+      "loss": 0.6049,
+      "step": 2734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2837781851403587,
+      "learning_rate": 1.8309544932829238e-05,
+      "loss": 0.6198,
+      "step": 2735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2952410964703984,
+      "learning_rate": 1.8308146770120893e-05,
+      "loss": 0.6702,
+      "step": 2736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3519706771227036,
+      "learning_rate": 1.8306748082876377e-05,
+      "loss": 0.6492,
+      "step": 2737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2180031664514333,
+      "learning_rate": 1.8305348871183998e-05,
+      "loss": 0.6082,
+      "step": 2738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2314869376762534,
+      "learning_rate": 1.8303949135132096e-05,
+      "loss": 0.6197,
+      "step": 2739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2433844982968294,
+      "learning_rate": 1.8302548874809042e-05,
+      "loss": 0.6288,
+      "step": 2740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.209795357841027,
+      "learning_rate": 1.8301148090303238e-05,
+      "loss": 0.625,
+      "step": 2741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0811275192303724,
+      "learning_rate": 1.829974678170313e-05,
+      "loss": 0.5735,
+      "step": 2742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1913268516224023,
+      "learning_rate": 1.8298344949097186e-05,
+      "loss": 0.5761,
+      "step": 2743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2868632195532392,
+      "learning_rate": 1.8296942592573912e-05,
+      "loss": 0.5971,
+      "step": 2744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2610298378187574,
+      "learning_rate": 1.8295539712221844e-05,
+      "loss": 0.5866,
+      "step": 2745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.14555682689655,
+      "learning_rate": 1.8294136308129556e-05,
+      "loss": 0.6317,
+      "step": 2746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1480329151219886,
+      "learning_rate": 1.829273238038565e-05,
+      "loss": 0.6175,
+      "step": 2747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3364857223610094,
+      "learning_rate": 1.8291327929078764e-05,
+      "loss": 0.6465,
+      "step": 2748
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.218675855679861,
+      "learning_rate": 1.828992295429757e-05,
+      "loss": 0.5915,
+      "step": 2749
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.069779256261487,
+      "learning_rate": 1.828851745613077e-05,
+      "loss": 0.5412,
+      "step": 2750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2703041892817974,
+      "learning_rate": 1.8287111434667096e-05,
+      "loss": 0.598,
+      "step": 2751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3296854613094637,
+      "learning_rate": 1.828570488999532e-05,
+      "loss": 0.6106,
+      "step": 2752
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2935112682245427,
+      "learning_rate": 1.8284297822204252e-05,
+      "loss": 0.6521,
+      "step": 2753
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.323029876230364,
+      "learning_rate": 1.8282890231382716e-05,
+      "loss": 0.6468,
+      "step": 2754
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2037016242300043,
+      "learning_rate": 1.828148211761959e-05,
+      "loss": 0.6173,
+      "step": 2755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2792445514258015,
+      "learning_rate": 1.828007348100377e-05,
+      "loss": 0.6499,
+      "step": 2756
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3389547084302056,
+      "learning_rate": 1.827866432162419e-05,
+      "loss": 0.6689,
+      "step": 2757
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.187141063050096,
+      "learning_rate": 1.8277254639569817e-05,
+      "loss": 0.6608,
+      "step": 2758
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3522860844756028,
+      "learning_rate": 1.827584443492966e-05,
+      "loss": 0.5747,
+      "step": 2759
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2429242138703271,
+      "learning_rate": 1.827443370779274e-05,
+      "loss": 0.669,
+      "step": 2760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1528596919165448,
+      "learning_rate": 1.827302245824813e-05,
+      "loss": 0.5587,
+      "step": 2761
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2648710194900268,
+      "learning_rate": 1.827161068638493e-05,
+      "loss": 0.6612,
+      "step": 2762
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3514441289498784,
+      "learning_rate": 1.8270198392292276e-05,
+      "loss": 0.7449,
+      "step": 2763
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1608384820495767,
+      "learning_rate": 1.8268785576059324e-05,
+      "loss": 0.5977,
+      "step": 2764
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3438503760174543,
+      "learning_rate": 1.8267372237775278e-05,
+      "loss": 0.6776,
+      "step": 2765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1346874111619902,
+      "learning_rate": 1.826595837752937e-05,
+      "loss": 0.613,
+      "step": 2766
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2276459786188239,
+      "learning_rate": 1.826454399541086e-05,
+      "loss": 0.6094,
+      "step": 2767
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1437264695378795,
+      "learning_rate": 1.826312909150905e-05,
+      "loss": 0.5423,
+      "step": 2768
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1743204667939153,
+      "learning_rate": 1.8261713665913268e-05,
+      "loss": 0.6243,
+      "step": 2769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3523323644801177,
+      "learning_rate": 1.8260297718712877e-05,
+      "loss": 0.6125,
+      "step": 2770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.23234212046008,
+      "learning_rate": 1.8258881249997275e-05,
+      "loss": 0.6303,
+      "step": 2771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2842448983116175,
+      "learning_rate": 1.825746425985589e-05,
+      "loss": 0.6576,
+      "step": 2772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4215279459532615,
+      "learning_rate": 1.825604674837818e-05,
+      "loss": 0.658,
+      "step": 2773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.340826047036285,
+      "learning_rate": 1.8254628715653647e-05,
+      "loss": 0.6525,
+      "step": 2774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1641832839464328,
+      "learning_rate": 1.825321016177181e-05,
+      "loss": 0.6072,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2655579820008127,
+      "learning_rate": 1.825179108682224e-05,
+      "loss": 0.663,
+      "step": 2776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2396023318760847,
+      "learning_rate": 1.8250371490894524e-05,
+      "loss": 0.6118,
+      "step": 2777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0523134426858805,
+      "learning_rate": 1.8248951374078288e-05,
+      "loss": 0.577,
+      "step": 2778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1822575243530766,
+      "learning_rate": 1.8247530736463194e-05,
+      "loss": 0.6171,
+      "step": 2779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3799827199351595,
+      "learning_rate": 1.8246109578138932e-05,
+      "loss": 0.6211,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.223821407802544,
+      "learning_rate": 1.8244687899195224e-05,
+      "loss": 0.651,
+      "step": 2781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1692815948648125,
+      "learning_rate": 1.824326569972184e-05,
+      "loss": 0.5911,
+      "step": 2782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.267942778767351,
+      "learning_rate": 1.8241842979808554e-05,
+      "loss": 0.6236,
+      "step": 2783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.221052537882822,
+      "learning_rate": 1.8240419739545206e-05,
+      "loss": 0.578,
+      "step": 2784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2035246841599823,
+      "learning_rate": 1.8238995979021642e-05,
+      "loss": 0.5699,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2816396911487473,
+      "learning_rate": 1.8237571698327752e-05,
+      "loss": 0.5972,
+      "step": 2786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2643952222970967,
+      "learning_rate": 1.8236146897553464e-05,
+      "loss": 0.6583,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3673255632716947,
+      "learning_rate": 1.8234721576788726e-05,
+      "loss": 0.6039,
+      "step": 2788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2348573442454402,
+      "learning_rate": 1.823329573612353e-05,
+      "loss": 0.6543,
+      "step": 2789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1745171325817638,
+      "learning_rate": 1.8231869375647897e-05,
+      "loss": 0.565,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2424715784917095,
+      "learning_rate": 1.823044249545188e-05,
+      "loss": 0.6605,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2302056409702453,
+      "learning_rate": 1.8229015095625567e-05,
+      "loss": 0.6465,
+      "step": 2792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.252995478150778,
+      "learning_rate": 1.822758717625907e-05,
+      "loss": 0.6635,
+      "step": 2793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2159450237244342,
+      "learning_rate": 1.8226158737442547e-05,
+      "loss": 0.6204,
+      "step": 2794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1720020988204702,
+      "learning_rate": 1.8224729779266183e-05,
+      "loss": 0.5526,
+      "step": 2795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1946464627716764,
+      "learning_rate": 1.8223300301820193e-05,
+      "loss": 0.6274,
+      "step": 2796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1868871311975935,
+      "learning_rate": 1.8221870305194828e-05,
+      "loss": 0.5913,
+      "step": 2797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3490003434443054,
+      "learning_rate": 1.8220439789480375e-05,
+      "loss": 0.6588,
+      "step": 2798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2190805622670569,
+      "learning_rate": 1.8219008754767144e-05,
+      "loss": 0.6279,
+      "step": 2799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3299475786447579,
+      "learning_rate": 1.8217577201145483e-05,
+      "loss": 0.6658,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1969145467016507,
+      "learning_rate": 1.821614512870578e-05,
+      "loss": 0.6126,
+      "step": 2801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1769735684599172,
+      "learning_rate": 1.821471253753844e-05,
+      "loss": 0.6276,
+      "step": 2802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4148715744779188,
+      "learning_rate": 1.821327942773392e-05,
+      "loss": 0.6855,
+      "step": 2803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3296521999097788,
+      "learning_rate": 1.8211845799382693e-05,
+      "loss": 0.6663,
+      "step": 2804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2144700239054251,
+      "learning_rate": 1.8210411652575275e-05,
+      "loss": 0.6591,
+      "step": 2805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3841934205573372,
+      "learning_rate": 1.8208976987402206e-05,
+      "loss": 0.6926,
+      "step": 2806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2289015219973962,
+      "learning_rate": 1.8207541803954068e-05,
+      "loss": 0.6341,
+      "step": 2807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2868690555800195,
+      "learning_rate": 1.8206106102321473e-05,
+      "loss": 0.5873,
+      "step": 2808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2454855460021577,
+      "learning_rate": 1.820466988259506e-05,
+      "loss": 0.6953,
+      "step": 2809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0982308292361256,
+      "learning_rate": 1.820323314486551e-05,
+      "loss": 0.5464,
+      "step": 2810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1006471875806687,
+      "learning_rate": 1.8201795889223525e-05,
+      "loss": 0.5914,
+      "step": 2811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2006455671781824,
+      "learning_rate": 1.8200358115759847e-05,
+      "loss": 0.5893,
+      "step": 2812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3924243301706416,
+      "learning_rate": 1.8198919824565258e-05,
+      "loss": 0.6627,
+      "step": 2813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4486124667564348,
+      "learning_rate": 1.8197481015730555e-05,
+      "loss": 0.6678,
+      "step": 2814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2937493402599183,
+      "learning_rate": 1.8196041689346585e-05,
+      "loss": 0.6645,
+      "step": 2815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3723180363975969,
+      "learning_rate": 1.819460184550422e-05,
+      "loss": 0.6615,
+      "step": 2816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2960667676916562,
+      "learning_rate": 1.8193161484294358e-05,
+      "loss": 0.6251,
+      "step": 2817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2350914240214055,
+      "learning_rate": 1.8191720605807937e-05,
+      "loss": 0.621,
+      "step": 2818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1791714177299912,
+      "learning_rate": 1.819027921013594e-05,
+      "loss": 0.5851,
+      "step": 2819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4171525271641732,
+      "learning_rate": 1.8188837297369353e-05,
+      "loss": 0.6497,
+      "step": 2820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.250284305665552,
+      "learning_rate": 1.818739486759922e-05,
+      "loss": 0.6881,
+      "step": 2821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3425361118396884,
+      "learning_rate": 1.8185951920916607e-05,
+      "loss": 0.655,
+      "step": 2822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2578889337856087,
+      "learning_rate": 1.8184508457412615e-05,
+      "loss": 0.6892,
+      "step": 2823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2068249177344736,
+      "learning_rate": 1.818306447717838e-05,
+      "loss": 0.6273,
+      "step": 2824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2200117182559702,
+      "learning_rate": 1.8181619980305065e-05,
+      "loss": 0.5969,
+      "step": 2825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2984507076909155,
+      "learning_rate": 1.818017496688387e-05,
+      "loss": 0.5938,
+      "step": 2826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2019157277396073,
+      "learning_rate": 1.817872943700602e-05,
+      "loss": 0.6362,
+      "step": 2827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2322419967809393,
+      "learning_rate": 1.817728339076279e-05,
+      "loss": 0.6147,
+      "step": 2828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.356758493475954,
+      "learning_rate": 1.8175836828245467e-05,
+      "loss": 0.6504,
+      "step": 2829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2897517817870001,
+      "learning_rate": 1.8174389749545384e-05,
+      "loss": 0.7105,
+      "step": 2830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2157061297255327,
+      "learning_rate": 1.81729421547539e-05,
+      "loss": 0.6304,
+      "step": 2831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2331171024493035,
+      "learning_rate": 1.8171494043962413e-05,
+      "loss": 0.6363,
+      "step": 2832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2129381941406479,
+      "learning_rate": 1.817004541726235e-05,
+      "loss": 0.6248,
+      "step": 2833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3303364462262077,
+      "learning_rate": 1.8168596274745165e-05,
+      "loss": 0.6512,
+      "step": 2834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1810256240319497,
+      "learning_rate": 1.8167146616502353e-05,
+      "loss": 0.6057,
+      "step": 2835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2015725560753407,
+      "learning_rate": 1.816569644262544e-05,
+      "loss": 0.6193,
+      "step": 2836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1208913268126142,
+      "learning_rate": 1.8164245753205977e-05,
+      "loss": 0.5856,
+      "step": 2837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1176206909308237,
+      "learning_rate": 1.816279454833556e-05,
+      "loss": 0.5683,
+      "step": 2838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3175210731746485,
+      "learning_rate": 1.8161342828105806e-05,
+      "loss": 0.6368,
+      "step": 2839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1667704479288525,
+      "learning_rate": 1.8159890592608378e-05,
+      "loss": 0.5324,
+      "step": 2840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0988289212991194,
+      "learning_rate": 1.8158437841934954e-05,
+      "loss": 0.5394,
+      "step": 2841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2842564549000774,
+      "learning_rate": 1.8156984576177258e-05,
+      "loss": 0.6238,
+      "step": 2842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2009700450512546,
+      "learning_rate": 1.815553079542704e-05,
+      "loss": 0.6667,
+      "step": 2843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.341760404443746,
+      "learning_rate": 1.8154076499776087e-05,
+      "loss": 0.6925,
+      "step": 2844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.303661834134969,
+      "learning_rate": 1.8152621689316216e-05,
+      "loss": 0.6543,
+      "step": 2845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1934411353340328,
+      "learning_rate": 1.815116636413927e-05,
+      "loss": 0.5741,
+      "step": 2846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2367567913207473,
+      "learning_rate": 1.8149710524337143e-05,
+      "loss": 0.5939,
+      "step": 2847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3560427459010786,
+      "learning_rate": 1.8148254170001743e-05,
+      "loss": 0.6495,
+      "step": 2848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.263134377740612,
+      "learning_rate": 1.8146797301225017e-05,
+      "loss": 0.6687,
+      "step": 2849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.173790701017271,
+      "learning_rate": 1.8145339918098944e-05,
+      "loss": 0.6114,
+      "step": 2850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1777002636475393,
+      "learning_rate": 1.8143882020715537e-05,
+      "loss": 0.5727,
+      "step": 2851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1990217115331079,
+      "learning_rate": 1.8142423609166845e-05,
+      "loss": 0.6531,
+      "step": 2852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2039315628866218,
+      "learning_rate": 1.8140964683544936e-05,
+      "loss": 0.626,
+      "step": 2853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2090947558373764,
+      "learning_rate": 1.813950524394193e-05,
+      "loss": 0.613,
+      "step": 2854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0763227907347506,
+      "learning_rate": 1.813804529044996e-05,
+      "loss": 0.5894,
+      "step": 2855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.103214621212828,
+      "learning_rate": 1.8136584823161205e-05,
+      "loss": 0.5624,
+      "step": 2856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.279610398872814,
+      "learning_rate": 1.8135123842167873e-05,
+      "loss": 0.5917,
+      "step": 2857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1060580243278069,
+      "learning_rate": 1.8133662347562197e-05,
+      "loss": 0.5864,
+      "step": 2858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.159203391728596,
+      "learning_rate": 1.8132200339436455e-05,
+      "loss": 0.5723,
+      "step": 2859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2723583961795326,
+      "learning_rate": 1.813073781788295e-05,
+      "loss": 0.6118,
+      "step": 2860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2273011140077157,
+      "learning_rate": 1.8129274782994016e-05,
+      "loss": 0.6806,
+      "step": 2861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2921420217350321,
+      "learning_rate": 1.8127811234862026e-05,
+      "loss": 0.6283,
+      "step": 2862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2323564369977353,
+      "learning_rate": 1.8126347173579373e-05,
+      "loss": 0.6112,
+      "step": 2863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.207304938931562,
+      "learning_rate": 1.8124882599238504e-05,
+      "loss": 0.6128,
+      "step": 2864
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1578066373977807,
+      "learning_rate": 1.812341751193187e-05,
+      "loss": 0.6047,
+      "step": 2865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4005586956898046,
+      "learning_rate": 1.812195191175198e-05,
+      "loss": 0.6594,
+      "step": 2866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.228123540125166,
+      "learning_rate": 1.8120485798791364e-05,
+      "loss": 0.6249,
+      "step": 2867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1568415391597346,
+      "learning_rate": 1.8119019173142583e-05,
+      "loss": 0.6069,
+      "step": 2868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1911612837300505,
+      "learning_rate": 1.811755203489823e-05,
+      "loss": 0.6288,
+      "step": 2869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.343495943759691,
+      "learning_rate": 1.811608438415094e-05,
+      "loss": 0.6332,
+      "step": 2870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2407355309741295,
+      "learning_rate": 1.8114616220993367e-05,
+      "loss": 0.6265,
+      "step": 2871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2773169185141842,
+      "learning_rate": 1.8113147545518207e-05,
+      "loss": 0.626,
+      "step": 2872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.252796050011533,
+      "learning_rate": 1.8111678357818184e-05,
+      "loss": 0.6127,
+      "step": 2873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.19824671414118,
+      "learning_rate": 1.8110208657986056e-05,
+      "loss": 0.6131,
+      "step": 2874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1883097949289403,
+      "learning_rate": 1.810873844611461e-05,
+      "loss": 0.5985,
+      "step": 2875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2814209521617395,
+      "learning_rate": 1.810726772229667e-05,
+      "loss": 0.6164,
+      "step": 2876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2923911841681872,
+      "learning_rate": 1.8105796486625095e-05,
+      "loss": 0.6373,
+      "step": 2877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2575262940617395,
+      "learning_rate": 1.8104324739192766e-05,
+      "loss": 0.6791,
+      "step": 2878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3457072884947838,
+      "learning_rate": 1.81028524800926e-05,
+      "loss": 0.6028,
+      "step": 2879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2840021859266266,
+      "learning_rate": 1.8101379709417556e-05,
+      "loss": 0.6638,
+      "step": 2880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2326226169186743,
+      "learning_rate": 1.809990642726061e-05,
+      "loss": 0.6518,
+      "step": 2881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1997694807368846,
+      "learning_rate": 1.809843263371478e-05,
+      "loss": 0.5913,
+      "step": 2882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2629495764488314,
+      "learning_rate": 1.809695832887312e-05,
+      "loss": 0.6485,
+      "step": 2883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.099901448516766,
+      "learning_rate": 1.8095483512828705e-05,
+      "loss": 0.6097,
+      "step": 2884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.231177328840679,
+      "learning_rate": 1.809400818567465e-05,
+      "loss": 0.6192,
+      "step": 2885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3685103637782334,
+      "learning_rate": 1.809253234750409e-05,
+      "loss": 0.641,
+      "step": 2886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1997628235875848,
+      "learning_rate": 1.809105599841022e-05,
+      "loss": 0.6401,
+      "step": 2887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0968343026199854,
+      "learning_rate": 1.8089579138486234e-05,
+      "loss": 0.5575,
+      "step": 2888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2138150879903193,
+      "learning_rate": 1.8088101767825385e-05,
+      "loss": 0.6557,
+      "step": 2889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2339204361674816,
+      "learning_rate": 1.8086623886520942e-05,
+      "loss": 0.6035,
+      "step": 2890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2712415237669306,
+      "learning_rate": 1.8085145494666208e-05,
+      "loss": 0.6331,
+      "step": 2891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2377801115733877,
+      "learning_rate": 1.808366659235453e-05,
+      "loss": 0.5811,
+      "step": 2892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3286606101312584,
+      "learning_rate": 1.8082187179679272e-05,
+      "loss": 0.6674,
+      "step": 2893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0994433316610714,
+      "learning_rate": 1.8080707256733837e-05,
+      "loss": 0.566,
+      "step": 2894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.223931181045098,
+      "learning_rate": 1.8079226823611665e-05,
+      "loss": 0.6318,
+      "step": 2895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.255745984975725,
+      "learning_rate": 1.807774588040622e-05,
+      "loss": 0.6282,
+      "step": 2896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.329411993834681,
+      "learning_rate": 1.8076264427210997e-05,
+      "loss": 0.6458,
+      "step": 2897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3125825129412299,
+      "learning_rate": 1.8074782464119536e-05,
+      "loss": 0.6372,
+      "step": 2898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.267016881705333,
+      "learning_rate": 1.8073299991225398e-05,
+      "loss": 0.5849,
+      "step": 2899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2533980435971066,
+      "learning_rate": 1.8071817008622177e-05,
+      "loss": 0.628,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2003904700783439,
+      "learning_rate": 1.8070333516403505e-05,
+      "loss": 0.5822,
+      "step": 2901
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2487166015160502,
+      "learning_rate": 1.806884951466304e-05,
+      "loss": 0.6062,
+      "step": 2902
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3581030584143734,
+      "learning_rate": 1.8067365003494475e-05,
+      "loss": 0.6601,
+      "step": 2903
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2631206459902808,
+      "learning_rate": 1.8065879982991536e-05,
+      "loss": 0.5929,
+      "step": 2904
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2821168990644012,
+      "learning_rate": 1.8064394453247977e-05,
+      "loss": 0.5959,
+      "step": 2905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2204427945062277,
+      "learning_rate": 1.8062908414357592e-05,
+      "loss": 0.5873,
+      "step": 2906
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.248558500725858,
+      "learning_rate": 1.80614218664142e-05,
+      "loss": 0.623,
+      "step": 2907
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2202886503382502,
+      "learning_rate": 1.8059934809511654e-05,
+      "loss": 0.5808,
+      "step": 2908
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2979357414415547,
+      "learning_rate": 1.805844724374384e-05,
+      "loss": 0.6001,
+      "step": 2909
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2011587746999037,
+      "learning_rate": 1.805695916920467e-05,
+      "loss": 0.5736,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3875532191612105,
+      "learning_rate": 1.8055470585988108e-05,
+      "loss": 0.6179,
+      "step": 2911
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2902609138404146,
+      "learning_rate": 1.805398149418812e-05,
+      "loss": 0.6515,
+      "step": 2912
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3347383884311403,
+      "learning_rate": 1.805249189389873e-05,
+      "loss": 0.6505,
+      "step": 2913
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3294166118671473,
+      "learning_rate": 1.805100178521398e-05,
+      "loss": 0.6658,
+      "step": 2914
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2300501035549545,
+      "learning_rate": 1.804951116822795e-05,
+      "loss": 0.6171,
+      "step": 2915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2054727115430035,
+      "learning_rate": 1.804802004303475e-05,
+      "loss": 0.6213,
+      "step": 2916
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3563128654571248,
+      "learning_rate": 1.8046528409728523e-05,
+      "loss": 0.6245,
+      "step": 2917
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3001366928567972,
+      "learning_rate": 1.8045036268403443e-05,
+      "loss": 0.6708,
+      "step": 2918
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1874836870378034,
+      "learning_rate": 1.8043543619153717e-05,
+      "loss": 0.6237,
+      "step": 2919
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.256463885722255,
+      "learning_rate": 1.804205046207358e-05,
+      "loss": 0.6283,
+      "step": 2920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2557813937595423,
+      "learning_rate": 1.8040556797257308e-05,
+      "loss": 0.5638,
+      "step": 2921
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2026769125661274,
+      "learning_rate": 1.80390626247992e-05,
+      "loss": 0.6409,
+      "step": 2922
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2871283152088284,
+      "learning_rate": 1.80375679447936e-05,
+      "loss": 0.6575,
+      "step": 2923
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2586342159232753,
+      "learning_rate": 1.803607275733486e-05,
+      "loss": 0.6328,
+      "step": 2924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1032225092944932,
+      "learning_rate": 1.8034577062517383e-05,
+      "loss": 0.5829,
+      "step": 2925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3104410370388593,
+      "learning_rate": 1.803308086043561e-05,
+      "loss": 0.6742,
+      "step": 2926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2051548863822021,
+      "learning_rate": 1.8031584151183995e-05,
+      "loss": 0.5936,
+      "step": 2927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2517796722089043,
+      "learning_rate": 1.8030086934857034e-05,
+      "loss": 0.6459,
+      "step": 2928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2320767507333181,
+      "learning_rate": 1.8028589211549256e-05,
+      "loss": 0.603,
+      "step": 2929
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2086792100061583,
+      "learning_rate": 1.8027090981355217e-05,
+      "loss": 0.6299,
+      "step": 2930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2700898804736247,
+      "learning_rate": 1.8025592244369514e-05,
+      "loss": 0.6605,
+      "step": 2931
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.206916927121023,
+      "learning_rate": 1.802409300068676e-05,
+      "loss": 0.614,
+      "step": 2932
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.168172608002909,
+      "learning_rate": 1.8022593250401625e-05,
+      "loss": 0.625,
+      "step": 2933
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1834166240034456,
+      "learning_rate": 1.8021092993608777e-05,
+      "loss": 0.5964,
+      "step": 2934
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2788494696389363,
+      "learning_rate": 1.801959223040295e-05,
+      "loss": 0.6234,
+      "step": 2935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.187926767860014,
+      "learning_rate": 1.801809096087889e-05,
+      "loss": 0.6049,
+      "step": 2936
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.267593407208171,
+      "learning_rate": 1.801658918513138e-05,
+      "loss": 0.6388,
+      "step": 2937
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.136411330349278,
+      "learning_rate": 1.801508690325523e-05,
+      "loss": 0.5344,
+      "step": 2938
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1600192724468503,
+      "learning_rate": 1.8013584115345297e-05,
+      "loss": 0.5305,
+      "step": 2939
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2332262899152213,
+      "learning_rate": 1.8012080821496454e-05,
+      "loss": 0.583,
+      "step": 2940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3254157313839539,
+      "learning_rate": 1.8010577021803608e-05,
+      "loss": 0.6546,
+      "step": 2941
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2185395866102802,
+      "learning_rate": 1.8009072716361707e-05,
+      "loss": 0.6275,
+      "step": 2942
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.271840316711964,
+      "learning_rate": 1.8007567905265727e-05,
+      "loss": 0.613,
+      "step": 2943
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.169847132567219,
+      "learning_rate": 1.8006062588610666e-05,
+      "loss": 0.5647,
+      "step": 2944
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.288062800386875,
+      "learning_rate": 1.8004556766491573e-05,
+      "loss": 0.6258,
+      "step": 2945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1649247630632729,
+      "learning_rate": 1.800305043900351e-05,
+      "loss": 0.5184,
+      "step": 2946
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1514814868520415,
+      "learning_rate": 1.8001543606241583e-05,
+      "loss": 0.6081,
+      "step": 2947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1776919127949836,
+      "learning_rate": 1.8000036268300927e-05,
+      "loss": 0.5703,
+      "step": 2948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3064339193428178,
+      "learning_rate": 1.7998528425276703e-05,
+      "loss": 0.6157,
+      "step": 2949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2702490083374884,
+      "learning_rate": 1.7997020077264118e-05,
+      "loss": 0.6398,
+      "step": 2950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2168450261809818,
+      "learning_rate": 1.799551122435839e-05,
+      "loss": 0.6088,
+      "step": 2951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.150140032328074,
+      "learning_rate": 1.7994001866654794e-05,
+      "loss": 0.5849,
+      "step": 2952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0685143506555084,
+      "learning_rate": 1.7992492004248615e-05,
+      "loss": 0.5281,
+      "step": 2953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2763959068563853,
+      "learning_rate": 1.7990981637235174e-05,
+      "loss": 0.607,
+      "step": 2954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1506804277702987,
+      "learning_rate": 1.798947076570984e-05,
+      "loss": 0.5811,
+      "step": 2955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1526404564556179,
+      "learning_rate": 1.7987959389767993e-05,
+      "loss": 0.6009,
+      "step": 2956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2005724892360312,
+      "learning_rate": 1.798644750950506e-05,
+      "loss": 0.6374,
+      "step": 2957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.14599489749152,
+      "learning_rate": 1.798493512501649e-05,
+      "loss": 0.5896,
+      "step": 2958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1765210431864042,
+      "learning_rate": 1.798342223639777e-05,
+      "loss": 0.6396,
+      "step": 2959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2504256954119262,
+      "learning_rate": 1.7981908843744413e-05,
+      "loss": 0.574,
+      "step": 2960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2331037614852411,
+      "learning_rate": 1.7980394947151972e-05,
+      "loss": 0.5922,
+      "step": 2961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.329365364258588,
+      "learning_rate": 1.797888054671602e-05,
+      "loss": 0.6719,
+      "step": 2962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2572771437637662,
+      "learning_rate": 1.7977365642532176e-05,
+      "loss": 0.6657,
+      "step": 2963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2879552999079014,
+      "learning_rate": 1.7975850234696084e-05,
+      "loss": 0.6882,
+      "step": 2964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2407461477170283,
+      "learning_rate": 1.7974334323303414e-05,
+      "loss": 0.6057,
+      "step": 2965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2215059859776771,
+      "learning_rate": 1.7972817908449875e-05,
+      "loss": 0.6278,
+      "step": 2966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.167622696761847,
+      "learning_rate": 1.7971300990231208e-05,
+      "loss": 0.6075,
+      "step": 2967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1778253169082262,
+      "learning_rate": 1.7969783568743185e-05,
+      "loss": 0.6018,
+      "step": 2968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2269449305997007,
+      "learning_rate": 1.7968265644081603e-05,
+      "loss": 0.5796,
+      "step": 2969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.31491470734766,
+      "learning_rate": 1.7966747216342305e-05,
+      "loss": 0.6621,
+      "step": 2970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.266386627101717,
+      "learning_rate": 1.7965228285621148e-05,
+      "loss": 0.6113,
+      "step": 2971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1610754719171927,
+      "learning_rate": 1.7963708852014034e-05,
+      "loss": 0.5741,
+      "step": 2972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2845976290844119,
+      "learning_rate": 1.7962188915616896e-05,
+      "loss": 0.6657,
+      "step": 2973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.130304124467323,
+      "learning_rate": 1.7960668476525688e-05,
+      "loss": 0.6081,
+      "step": 2974
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2494207470577738,
+      "learning_rate": 1.795914753483641e-05,
+      "loss": 0.5918,
+      "step": 2975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2301678486406489,
+      "learning_rate": 1.7957626090645085e-05,
+      "loss": 0.619,
+      "step": 2976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1322886867302662,
+      "learning_rate": 1.7956104144047768e-05,
+      "loss": 0.5837,
+      "step": 2977
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1486553063449865,
+      "learning_rate": 1.795458169514055e-05,
+      "loss": 0.5828,
+      "step": 2978
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.160789137302683,
+      "learning_rate": 1.7953058744019546e-05,
+      "loss": 0.5474,
+      "step": 2979
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2342475632181233,
+      "learning_rate": 1.7951535290780913e-05,
+      "loss": 0.5618,
+      "step": 2980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.270766376336555,
+      "learning_rate": 1.7950011335520832e-05,
+      "loss": 0.5931,
+      "step": 2981
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2507334941298718,
+      "learning_rate": 1.7948486878335522e-05,
+      "loss": 0.5865,
+      "step": 2982
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5868925779795653,
+      "learning_rate": 1.7946961919321225e-05,
+      "loss": 0.6711,
+      "step": 2983
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1474935064360614,
+      "learning_rate": 1.7945436458574216e-05,
+      "loss": 0.5709,
+      "step": 2984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1264940513524888,
+      "learning_rate": 1.7943910496190816e-05,
+      "loss": 0.584,
+      "step": 2985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2317083986789328,
+      "learning_rate": 1.794238403226736e-05,
+      "loss": 0.6154,
+      "step": 2986
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1544663874415577,
+      "learning_rate": 1.7940857066900223e-05,
+      "loss": 0.5924,
+      "step": 2987
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.277005445377626,
+      "learning_rate": 1.7939329600185807e-05,
+      "loss": 0.6846,
+      "step": 2988
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1453012213470386,
+      "learning_rate": 1.7937801632220556e-05,
+      "loss": 0.5462,
+      "step": 2989
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2065896025614506,
+      "learning_rate": 1.793627316310093e-05,
+      "loss": 0.589,
+      "step": 2990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1848137236710226,
+      "learning_rate": 1.7934744192923436e-05,
+      "loss": 0.6013,
+      "step": 2991
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2619593246270835,
+      "learning_rate": 1.7933214721784602e-05,
+      "loss": 0.6333,
+      "step": 2992
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2150648596606142,
+      "learning_rate": 1.7931684749780994e-05,
+      "loss": 0.6105,
+      "step": 2993
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2555330840400099,
+      "learning_rate": 1.7930154277009207e-05,
+      "loss": 0.654,
+      "step": 2994
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1743399064533986,
+      "learning_rate": 1.792862330356586e-05,
+      "loss": 0.6053,
+      "step": 2995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.164000387902867,
+      "learning_rate": 1.7927091829547624e-05,
+      "loss": 0.541,
+      "step": 2996
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2705627476443653,
+      "learning_rate": 1.792555985505118e-05,
+      "loss": 0.6016,
+      "step": 2997
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2389304688427403,
+      "learning_rate": 1.7924027380173253e-05,
+      "loss": 0.5954,
+      "step": 2998
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1266852577827493,
+      "learning_rate": 1.7922494405010593e-05,
+      "loss": 0.691,
+      "step": 2999
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2202408304305663,
+      "learning_rate": 1.792096092965999e-05,
+      "loss": 0.6233,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2500432960641452,
+      "learning_rate": 1.7919426954218252e-05,
+      "loss": 0.6002,
+      "step": 3001
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2753983828905995,
+      "learning_rate": 1.7917892478782234e-05,
+      "loss": 0.6494,
+      "step": 3002
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1316247962000014,
+      "learning_rate": 1.791635750344881e-05,
+      "loss": 0.5839,
+      "step": 3003
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3513382742223259,
+      "learning_rate": 1.79148220283149e-05,
+      "loss": 0.6534,
+      "step": 3004
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3354423006704952,
+      "learning_rate": 1.7913286053477434e-05,
+      "loss": 0.689,
+      "step": 3005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.293549790301838,
+      "learning_rate": 1.7911749579033394e-05,
+      "loss": 0.7231,
+      "step": 3006
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1181166193196503,
+      "learning_rate": 1.7910212605079788e-05,
+      "loss": 0.5189,
+      "step": 3007
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2468709883469986,
+      "learning_rate": 1.7908675131713642e-05,
+      "loss": 0.5978,
+      "step": 3008
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1993187480574705,
+      "learning_rate": 1.7907137159032036e-05,
+      "loss": 0.6734,
+      "step": 3009
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0643164870731645,
+      "learning_rate": 1.7905598687132064e-05,
+      "loss": 0.5736,
+      "step": 3010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2758891371306944,
+      "learning_rate": 1.790405971611086e-05,
+      "loss": 0.589,
+      "step": 3011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2234611421251935,
+      "learning_rate": 1.790252024606559e-05,
+      "loss": 0.6108,
+      "step": 3012
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1952096234100136,
+      "learning_rate": 1.7900980277093438e-05,
+      "loss": 0.6469,
+      "step": 3013
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3264603448541694,
+      "learning_rate": 1.7899439809291643e-05,
+      "loss": 0.5981,
+      "step": 3014
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1788633891126337,
+      "learning_rate": 1.7897898842757455e-05,
+      "loss": 0.577,
+      "step": 3015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1995112655702147,
+      "learning_rate": 1.7896357377588165e-05,
+      "loss": 0.5554,
+      "step": 3016
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3007924763879526,
+      "learning_rate": 1.7894815413881096e-05,
+      "loss": 0.693,
+      "step": 3017
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.31782970950039,
+      "learning_rate": 1.7893272951733596e-05,
+      "loss": 0.6413,
+      "step": 3018
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.345985061942385,
+      "learning_rate": 1.789172999124305e-05,
+      "loss": 0.6494,
+      "step": 3019
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.153175855306034,
+      "learning_rate": 1.7890186532506875e-05,
+      "loss": 0.5905,
+      "step": 3020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2546485771327105,
+      "learning_rate": 1.788864257562252e-05,
+      "loss": 0.6302,
+      "step": 3021
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2414322962588247,
+      "learning_rate": 1.788709812068745e-05,
+      "loss": 0.6255,
+      "step": 3022
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1241496899291201,
+      "learning_rate": 1.788555316779919e-05,
+      "loss": 0.6019,
+      "step": 3023
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2547422575254765,
+      "learning_rate": 1.7884007717055273e-05,
+      "loss": 0.6121,
+      "step": 3024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2316432131356856,
+      "learning_rate": 1.7882461768553275e-05,
+      "loss": 0.5705,
+      "step": 3025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.209348460455779,
+      "learning_rate": 1.7880915322390794e-05,
+      "loss": 0.6658,
+      "step": 3026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.331915484009588,
+      "learning_rate": 1.787936837866547e-05,
+      "loss": 0.6434,
+      "step": 3027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3494130501057997,
+      "learning_rate": 1.7877820937474966e-05,
+      "loss": 0.6342,
+      "step": 3028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2896999285879898,
+      "learning_rate": 1.7876272998916987e-05,
+      "loss": 0.6691,
+      "step": 3029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2085400875830674,
+      "learning_rate": 1.787472456308925e-05,
+      "loss": 0.5967,
+      "step": 3030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1905624653571254,
+      "learning_rate": 1.787317563008953e-05,
+      "loss": 0.6734,
+      "step": 3031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2800279697700183,
+      "learning_rate": 1.7871626200015607e-05,
+      "loss": 0.7052,
+      "step": 3032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1663746752237185,
+      "learning_rate": 1.7870076272965313e-05,
+      "loss": 0.5192,
+      "step": 3033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2075594638217475,
+      "learning_rate": 1.78685258490365e-05,
+      "loss": 0.5684,
+      "step": 3034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1376299794310991,
+      "learning_rate": 1.7866974928327052e-05,
+      "loss": 0.5841,
+      "step": 3035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2071017149075514,
+      "learning_rate": 1.7865423510934888e-05,
+      "loss": 0.5715,
+      "step": 3036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0675487616165151,
+      "learning_rate": 1.7863871596957963e-05,
+      "loss": 0.5698,
+      "step": 3037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3310970039386116,
+      "learning_rate": 1.7862319186494245e-05,
+      "loss": 0.6288,
+      "step": 3038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.137937174616564,
+      "learning_rate": 1.7860766279641758e-05,
+      "loss": 0.6085,
+      "step": 3039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1180286575356846,
+      "learning_rate": 1.7859212876498536e-05,
+      "loss": 0.5583,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.184991143833337,
+      "learning_rate": 1.7857658977162655e-05,
+      "loss": 0.5962,
+      "step": 3041
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1759104610176012,
+      "learning_rate": 1.7856104581732227e-05,
+      "loss": 0.6277,
+      "step": 3042
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.216858006590385,
+      "learning_rate": 1.785454969030538e-05,
+      "loss": 0.6217,
+      "step": 3043
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.218961110537572,
+      "learning_rate": 1.785299430298029e-05,
+      "loss": 0.5992,
+      "step": 3044
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1718555703141884,
+      "learning_rate": 1.7851438419855157e-05,
+      "loss": 0.5835,
+      "step": 3045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0695880749405033,
+      "learning_rate": 1.7849882041028203e-05,
+      "loss": 0.539,
+      "step": 3046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0857060137680985,
+      "learning_rate": 1.7848325166597698e-05,
+      "loss": 0.555,
+      "step": 3047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.255971900561838,
+      "learning_rate": 1.7846767796661934e-05,
+      "loss": 0.6104,
+      "step": 3048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.21615372954044,
+      "learning_rate": 1.7845209931319232e-05,
+      "loss": 0.657,
+      "step": 3049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1972051857259582,
+      "learning_rate": 1.7843651570667953e-05,
+      "loss": 0.5963,
+      "step": 3050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1667782128460005,
+      "learning_rate": 1.784209271480648e-05,
+      "loss": 0.6278,
+      "step": 3051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2684358536737081,
+      "learning_rate": 1.7840533363833238e-05,
+      "loss": 0.5818,
+      "step": 3052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3054852359627946,
+      "learning_rate": 1.783897351784667e-05,
+      "loss": 0.6701,
+      "step": 3053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2808090474629414,
+      "learning_rate": 1.7837413176945263e-05,
+      "loss": 0.5995,
+      "step": 3054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.26955472484306,
+      "learning_rate": 1.7835852341227523e-05,
+      "loss": 0.6308,
+      "step": 3055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3128604848001724,
+      "learning_rate": 1.7834291010791998e-05,
+      "loss": 0.6654,
+      "step": 3056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.185201025703545,
+      "learning_rate": 1.783272918573726e-05,
+      "loss": 0.6088,
+      "step": 3057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1497198198923226,
+      "learning_rate": 1.7831166866161923e-05,
+      "loss": 0.625,
+      "step": 3058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.135485055012956,
+      "learning_rate": 1.7829604052164616e-05,
+      "loss": 0.5678,
+      "step": 3059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0077649482331934,
+      "learning_rate": 1.7828040743844008e-05,
+      "loss": 0.5576,
+      "step": 3060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1692780265785083,
+      "learning_rate": 1.7826476941298797e-05,
+      "loss": 0.5841,
+      "step": 3061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2531621513662476,
+      "learning_rate": 1.782491264462772e-05,
+      "loss": 0.5952,
+      "step": 3062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1774960809792259,
+      "learning_rate": 1.782334785392954e-05,
+      "loss": 0.6002,
+      "step": 3063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3198974227785316,
+      "learning_rate": 1.7821782569303045e-05,
+      "loss": 0.6301,
+      "step": 3064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2412421509954603,
+      "learning_rate": 1.782021679084706e-05,
+      "loss": 0.6383,
+      "step": 3065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2281711016439567,
+      "learning_rate": 1.781865051866044e-05,
+      "loss": 0.5641,
+      "step": 3066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2609974124860723,
+      "learning_rate": 1.781708375284208e-05,
+      "loss": 0.6578,
+      "step": 3067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2377115377091337,
+      "learning_rate": 1.7815516493490888e-05,
+      "loss": 0.5483,
+      "step": 3068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1886722901914275,
+      "learning_rate": 1.7813948740705816e-05,
+      "loss": 0.5809,
+      "step": 3069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3063399763010588,
+      "learning_rate": 1.7812380494585846e-05,
+      "loss": 0.6425,
+      "step": 3070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2197979554172487,
+      "learning_rate": 1.781081175522999e-05,
+      "loss": 0.6027,
+      "step": 3071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3226034840008223,
+      "learning_rate": 1.780924252273729e-05,
+      "loss": 0.6882,
+      "step": 3072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2641265381387161,
+      "learning_rate": 1.780767279720682e-05,
+      "loss": 0.593,
+      "step": 3073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1856570247654552,
+      "learning_rate": 1.780610257873768e-05,
+      "loss": 0.5957,
+      "step": 3074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2788455545590587,
+      "learning_rate": 1.7804531867429013e-05,
+      "loss": 0.5838,
+      "step": 3075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0892644183290483,
+      "learning_rate": 1.780296066337998e-05,
+      "loss": 0.5805,
+      "step": 3076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2363920029564532,
+      "learning_rate": 1.7801388966689784e-05,
+      "loss": 0.6094,
+      "step": 3077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1601366754286944,
+      "learning_rate": 1.7799816777457653e-05,
+      "loss": 0.5863,
+      "step": 3078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1786872210325052,
+      "learning_rate": 1.7798244095782847e-05,
+      "loss": 0.5889,
+      "step": 3079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2907588547575215,
+      "learning_rate": 1.779667092176466e-05,
+      "loss": 0.6372,
+      "step": 3080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2001794700725412,
+      "learning_rate": 1.779509725550241e-05,
+      "loss": 0.604,
+      "step": 3081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2630275398417172,
+      "learning_rate": 1.7793523097095452e-05,
+      "loss": 0.5937,
+      "step": 3082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2784739868954929,
+      "learning_rate": 1.7791948446643173e-05,
+      "loss": 0.563,
+      "step": 3083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1527263975885382,
+      "learning_rate": 1.7790373304244986e-05,
+      "loss": 0.5368,
+      "step": 3084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2456382469012621,
+      "learning_rate": 1.778879767000034e-05,
+      "loss": 0.6421,
+      "step": 3085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2194838637457066,
+      "learning_rate": 1.7787221544008715e-05,
+      "loss": 0.5656,
+      "step": 3086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.273948192582068,
+      "learning_rate": 1.7785644926369616e-05,
+      "loss": 0.6647,
+      "step": 3087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2184152387927174,
+      "learning_rate": 1.778406781718258e-05,
+      "loss": 0.6336,
+      "step": 3088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3443164962034504,
+      "learning_rate": 1.7782490216547182e-05,
+      "loss": 0.5989,
+      "step": 3089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2599534005676034,
+      "learning_rate": 1.778091212456303e-05,
+      "loss": 0.6655,
+      "step": 3090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2018116805466454,
+      "learning_rate": 1.7779333541329745e-05,
+      "loss": 0.6561,
+      "step": 3091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.184220201748367,
+      "learning_rate": 1.7777754466947002e-05,
+      "loss": 0.5243,
+      "step": 3092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1420561290916702,
+      "learning_rate": 1.7776174901514493e-05,
+      "loss": 0.5828,
+      "step": 3093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1784786068087671,
+      "learning_rate": 1.777459484513194e-05,
+      "loss": 0.5923,
+      "step": 3094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2604503574545003,
+      "learning_rate": 1.77730142978991e-05,
+      "loss": 0.5698,
+      "step": 3095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2245177214910639,
+      "learning_rate": 1.7771433259915767e-05,
+      "loss": 0.6243,
+      "step": 3096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1310737156374882,
+      "learning_rate": 1.7769851731281758e-05,
+      "loss": 0.5655,
+      "step": 3097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2791102149427254,
+      "learning_rate": 1.7768269712096922e-05,
+      "loss": 0.5306,
+      "step": 3098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3072608650297313,
+      "learning_rate": 1.7766687202461137e-05,
+      "loss": 0.6488,
+      "step": 3099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3758335188078605,
+      "learning_rate": 1.776510420247432e-05,
+      "loss": 0.7007,
+      "step": 3100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2715708150940994,
+      "learning_rate": 1.7763520712236414e-05,
+      "loss": 0.6378,
+      "step": 3101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2376673286673203,
+      "learning_rate": 1.776193673184739e-05,
+      "loss": 0.6697,
+      "step": 3102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.293950425623742,
+      "learning_rate": 1.776035226140725e-05,
+      "loss": 0.6214,
+      "step": 3103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1915457221770092,
+      "learning_rate": 1.7758767301016034e-05,
+      "loss": 0.6003,
+      "step": 3104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3364921890612806,
+      "learning_rate": 1.7757181850773812e-05,
+      "loss": 0.6393,
+      "step": 3105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.243742012113291,
+      "learning_rate": 1.7755595910780677e-05,
+      "loss": 0.625,
+      "step": 3106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1462533643334123,
+      "learning_rate": 1.7754009481136758e-05,
+      "loss": 0.5895,
+      "step": 3107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2663942518848426,
+      "learning_rate": 1.7752422561942216e-05,
+      "loss": 0.6783,
+      "step": 3108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2311750534406798,
+      "learning_rate": 1.7750835153297238e-05,
+      "loss": 0.6015,
+      "step": 3109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1748801291985216,
+      "learning_rate": 1.774924725530205e-05,
+      "loss": 0.579,
+      "step": 3110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.225639919596761,
+      "learning_rate": 1.77476588680569e-05,
+      "loss": 0.6048,
+      "step": 3111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0920129879434115,
+      "learning_rate": 1.7746069991662076e-05,
+      "loss": 0.588,
+      "step": 3112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2202342849654233,
+      "learning_rate": 1.7744480626217886e-05,
+      "loss": 0.6283,
+      "step": 3113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2688035963041178,
+      "learning_rate": 1.774289077182468e-05,
+      "loss": 0.6084,
+      "step": 3114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2619185629018865,
+      "learning_rate": 1.7741300428582827e-05,
+      "loss": 0.5455,
+      "step": 3115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.106099141066414,
+      "learning_rate": 1.7739709596592742e-05,
+      "loss": 0.5378,
+      "step": 3116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2088591922171792,
+      "learning_rate": 1.773811827595486e-05,
+      "loss": 0.5974,
+      "step": 3117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2637047977079623,
+      "learning_rate": 1.7736526466769645e-05,
+      "loss": 0.6625,
+      "step": 3118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3631647710582162,
+      "learning_rate": 1.77349341691376e-05,
+      "loss": 0.6746,
+      "step": 3119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2091894022151295,
+      "learning_rate": 1.7733341383159254e-05,
+      "loss": 0.5602,
+      "step": 3120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3337131743991497,
+      "learning_rate": 1.773174810893517e-05,
+      "loss": 0.6713,
+      "step": 3121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1644816316111655,
+      "learning_rate": 1.7730154346565932e-05,
+      "loss": 0.6071,
+      "step": 3122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2607845000857336,
+      "learning_rate": 1.772856009615217e-05,
+      "loss": 0.6426,
+      "step": 3123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1358324508682223,
+      "learning_rate": 1.7726965357794536e-05,
+      "loss": 0.5585,
+      "step": 3124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2403745557005192,
+      "learning_rate": 1.7725370131593713e-05,
+      "loss": 0.6094,
+      "step": 3125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.174291027532417,
+      "learning_rate": 1.7723774417650415e-05,
+      "loss": 0.602,
+      "step": 3126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2481000288985398,
+      "learning_rate": 1.772217821606539e-05,
+      "loss": 0.6371,
+      "step": 3127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3395270148383536,
+      "learning_rate": 1.7720581526939412e-05,
+      "loss": 0.6196,
+      "step": 3128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2598934138341693,
+      "learning_rate": 1.771898435037329e-05,
+      "loss": 0.5804,
+      "step": 3129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3108867539960758,
+      "learning_rate": 1.7717386686467856e-05,
+      "loss": 0.5992,
+      "step": 3130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.355990526543217,
+      "learning_rate": 1.771578853532399e-05,
+      "loss": 0.6602,
+      "step": 3131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2479598562741128,
+      "learning_rate": 1.7714189897042583e-05,
+      "loss": 0.6633,
+      "step": 3132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3442980070248483,
+      "learning_rate": 1.7712590771724572e-05,
+      "loss": 0.6271,
+      "step": 3133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2984706300624225,
+      "learning_rate": 1.771099115947091e-05,
+      "loss": 0.6151,
+      "step": 3134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2461555967857512,
+      "learning_rate": 1.7709391060382592e-05,
+      "loss": 0.6417,
+      "step": 3135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3102342030781724,
+      "learning_rate": 1.7707790474560644e-05,
+      "loss": 0.632,
+      "step": 3136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1246453361949431,
+      "learning_rate": 1.7706189402106113e-05,
+      "loss": 0.5425,
+      "step": 3137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2233872346064572,
+      "learning_rate": 1.770458784312009e-05,
+      "loss": 0.6269,
+      "step": 3138
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2542860460125764,
+      "learning_rate": 1.7702985797703682e-05,
+      "loss": 0.6131,
+      "step": 3139
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2888452982702112,
+      "learning_rate": 1.7701383265958042e-05,
+      "loss": 0.6223,
+      "step": 3140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1345702641678044,
+      "learning_rate": 1.7699780247984343e-05,
+      "loss": 0.6009,
+      "step": 3141
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.23611269937668,
+      "learning_rate": 1.7698176743883786e-05,
+      "loss": 0.6078,
+      "step": 3142
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2875427072349082,
+      "learning_rate": 1.769657275375762e-05,
+      "loss": 0.681,
+      "step": 3143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1695176383062844,
+      "learning_rate": 1.7694968277707102e-05,
+      "loss": 0.582,
+      "step": 3144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1835082876240008,
+      "learning_rate": 1.769336331583354e-05,
+      "loss": 0.6119,
+      "step": 3145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.199357413063099,
+      "learning_rate": 1.7691757868238256e-05,
+      "loss": 0.6433,
+      "step": 3146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1096665979692646,
+      "learning_rate": 1.7690151935022616e-05,
+      "loss": 0.5449,
+      "step": 3147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2888041844340898,
+      "learning_rate": 1.7688545516288006e-05,
+      "loss": 0.6345,
+      "step": 3148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.388267146926027,
+      "learning_rate": 1.768693861213585e-05,
+      "loss": 0.6499,
+      "step": 3149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.085520053462167,
+      "learning_rate": 1.76853312226676e-05,
+      "loss": 0.55,
+      "step": 3150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1869019457995977,
+      "learning_rate": 1.768372334798474e-05,
+      "loss": 0.59,
+      "step": 3151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2856650229507738,
+      "learning_rate": 1.7682114988188782e-05,
+      "loss": 0.5742,
+      "step": 3152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.292939661165538,
+      "learning_rate": 1.768050614338127e-05,
+      "loss": 0.6041,
+      "step": 3153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0942013490462523,
+      "learning_rate": 1.767889681366378e-05,
+      "loss": 0.5807,
+      "step": 3154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.208738434508418,
+      "learning_rate": 1.7677286999137916e-05,
+      "loss": 0.5863,
+      "step": 3155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2480825022118012,
+      "learning_rate": 1.767567669990531e-05,
+      "loss": 0.5904,
+      "step": 3156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2206445786741504,
+      "learning_rate": 1.767406591606764e-05,
+      "loss": 0.5646,
+      "step": 3157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.236043791896361,
+      "learning_rate": 1.767245464772659e-05,
+      "loss": 0.5646,
+      "step": 3158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.17166435891725,
+      "learning_rate": 1.7670842894983898e-05,
+      "loss": 0.6143,
+      "step": 3159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1660721092664443,
+      "learning_rate": 1.7669230657941314e-05,
+      "loss": 0.6108,
+      "step": 3160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3440181109848686,
+      "learning_rate": 1.7667617936700634e-05,
+      "loss": 0.6317,
+      "step": 3161
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2503476136379494,
+      "learning_rate": 1.766600473136367e-05,
+      "loss": 0.6245,
+      "step": 3162
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2463578090888154,
+      "learning_rate": 1.7664391042032277e-05,
+      "loss": 0.6237,
+      "step": 3163
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3054172964588686,
+      "learning_rate": 1.766277686880834e-05,
+      "loss": 0.6322,
+      "step": 3164
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2412940116831916,
+      "learning_rate": 1.7661162211793757e-05,
+      "loss": 0.6475,
+      "step": 3165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.227847355721397,
+      "learning_rate": 1.765954707109048e-05,
+      "loss": 0.6049,
+      "step": 3166
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2741661093648053,
+      "learning_rate": 1.7657931446800475e-05,
+      "loss": 0.6355,
+      "step": 3167
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2351243364239475,
+      "learning_rate": 1.7656315339025754e-05,
+      "loss": 0.6611,
+      "step": 3168
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.293660650070407,
+      "learning_rate": 1.765469874786834e-05,
+      "loss": 0.5834,
+      "step": 3169
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.28034629585367,
+      "learning_rate": 1.7653081673430302e-05,
+      "loss": 0.6123,
+      "step": 3170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.161525137082055,
+      "learning_rate": 1.7651464115813738e-05,
+      "loss": 0.5789,
+      "step": 3171
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3447536446386683,
+      "learning_rate": 1.764984607512076e-05,
+      "loss": 0.6256,
+      "step": 3172
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.189310450523712,
+      "learning_rate": 1.7648227551453535e-05,
+      "loss": 0.5704,
+      "step": 3173
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1945091490013942,
+      "learning_rate": 1.7646608544914245e-05,
+      "loss": 0.6375,
+      "step": 3174
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1590937620355648,
+      "learning_rate": 1.764498905560511e-05,
+      "loss": 0.6081,
+      "step": 3175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2789708311708203,
+      "learning_rate": 1.764336908362837e-05,
+      "loss": 0.6274,
+      "step": 3176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.224904050767292,
+      "learning_rate": 1.76417486290863e-05,
+      "loss": 0.5833,
+      "step": 3177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2411030768253288,
+      "learning_rate": 1.764012769208122e-05,
+      "loss": 0.6002,
+      "step": 3178
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1292730349356888,
+      "learning_rate": 1.7638506272715458e-05,
+      "loss": 0.5581,
+      "step": 3179
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.231072656087877,
+      "learning_rate": 1.7636884371091385e-05,
+      "loss": 0.6251,
+      "step": 3180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1978900519468991,
+      "learning_rate": 1.76352619873114e-05,
+      "loss": 0.6132,
+      "step": 3181
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.121095185494916,
+      "learning_rate": 1.7633639121477935e-05,
+      "loss": 0.5977,
+      "step": 3182
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1506382104217876,
+      "learning_rate": 1.7632015773693444e-05,
+      "loss": 0.6299,
+      "step": 3183
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1663758505802897,
+      "learning_rate": 1.7630391944060424e-05,
+      "loss": 0.6218,
+      "step": 3184
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2367705266238813,
+      "learning_rate": 1.7628767632681393e-05,
+      "loss": 0.622,
+      "step": 3185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2750108045232404,
+      "learning_rate": 1.7627142839658903e-05,
+      "loss": 0.6443,
+      "step": 3186
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.267974838458455,
+      "learning_rate": 1.7625517565095536e-05,
+      "loss": 0.6656,
+      "step": 3187
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2187938682290484,
+      "learning_rate": 1.76238918090939e-05,
+      "loss": 0.613,
+      "step": 3188
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2599129524134998,
+      "learning_rate": 1.762226557175664e-05,
+      "loss": 0.6017,
+      "step": 3189
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1959538453328042,
+      "learning_rate": 1.762063885318643e-05,
+      "loss": 0.581,
+      "step": 3190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2414900063071381,
+      "learning_rate": 1.7619011653485968e-05,
+      "loss": 0.5808,
+      "step": 3191
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2167196233973314,
+      "learning_rate": 1.7617383972758e-05,
+      "loss": 0.6234,
+      "step": 3192
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1507774441015806,
+      "learning_rate": 1.761575581110527e-05,
+      "loss": 0.5854,
+      "step": 3193
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.314947842916153,
+      "learning_rate": 1.7614127168630594e-05,
+      "loss": 0.6175,
+      "step": 3194
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.123485446378992,
+      "learning_rate": 1.7612498045436778e-05,
+      "loss": 0.5849,
+      "step": 3195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1255848212222526,
+      "learning_rate": 1.761086844162669e-05,
+      "loss": 0.5876,
+      "step": 3196
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1940461475457604,
+      "learning_rate": 1.7609238357303208e-05,
+      "loss": 0.6501,
+      "step": 3197
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.233503347784292,
+      "learning_rate": 1.760760779256925e-05,
+      "loss": 0.6747,
+      "step": 3198
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3401509369349782,
+      "learning_rate": 1.7605976747527763e-05,
+      "loss": 0.6246,
+      "step": 3199
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2139890061843788,
+      "learning_rate": 1.760434522228172e-05,
+      "loss": 0.5566,
+      "step": 3200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2199347190901457,
+      "learning_rate": 1.7602713216934132e-05,
+      "loss": 0.6276,
+      "step": 3201
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1978119790593345,
+      "learning_rate": 1.7601080731588034e-05,
+      "loss": 0.611,
+      "step": 3202
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2223582041599093,
+      "learning_rate": 1.7599447766346494e-05,
+      "loss": 0.6594,
+      "step": 3203
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2071774094861256,
+      "learning_rate": 1.7597814321312605e-05,
+      "loss": 0.6109,
+      "step": 3204
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1836781487416035,
+      "learning_rate": 1.75961803965895e-05,
+      "loss": 0.5251,
+      "step": 3205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3086132902138354,
+      "learning_rate": 1.7594545992280336e-05,
+      "loss": 0.5832,
+      "step": 3206
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1603592717457671,
+      "learning_rate": 1.75929111084883e-05,
+      "loss": 0.6271,
+      "step": 3207
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.247102621999723,
+      "learning_rate": 1.7591275745316613e-05,
+      "loss": 0.6509,
+      "step": 3208
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1599715885431923,
+      "learning_rate": 1.7589639902868524e-05,
+      "loss": 0.5862,
+      "step": 3209
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2619552626834056,
+      "learning_rate": 1.7588003581247307e-05,
+      "loss": 0.6285,
+      "step": 3210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.265983719086781,
+      "learning_rate": 1.7586366780556275e-05,
+      "loss": 0.6405,
+      "step": 3211
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1950286827475802,
+      "learning_rate": 1.7584729500898768e-05,
+      "loss": 0.5574,
+      "step": 3212
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.270312447448204,
+      "learning_rate": 1.7583091742378157e-05,
+      "loss": 0.6536,
+      "step": 3213
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1762339078671982,
+      "learning_rate": 1.7581453505097838e-05,
+      "loss": 0.6238,
+      "step": 3214
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.248898402227216,
+      "learning_rate": 1.7579814789161246e-05,
+      "loss": 0.6491,
+      "step": 3215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2109732837927323,
+      "learning_rate": 1.757817559467184e-05,
+      "loss": 0.5537,
+      "step": 3216
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1729884135299256,
+      "learning_rate": 1.757653592173311e-05,
+      "loss": 0.6663,
+      "step": 3217
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2556395627549655,
+      "learning_rate": 1.7574895770448578e-05,
+      "loss": 0.5797,
+      "step": 3218
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.195463881536923,
+      "learning_rate": 1.7573255140921793e-05,
+      "loss": 0.6031,
+      "step": 3219
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2584785451321943,
+      "learning_rate": 1.7571614033256338e-05,
+      "loss": 0.6481,
+      "step": 3220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1928259223820412,
+      "learning_rate": 1.7569972447555827e-05,
+      "loss": 0.5866,
+      "step": 3221
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2776793059584677,
+      "learning_rate": 1.7568330383923902e-05,
+      "loss": 0.6228,
+      "step": 3222
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1727822924128048,
+      "learning_rate": 1.7566687842464224e-05,
+      "loss": 0.5913,
+      "step": 3223
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2486822812159255,
+      "learning_rate": 1.756504482328051e-05,
+      "loss": 0.6199,
+      "step": 3224
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.269318689011064,
+      "learning_rate": 1.7563401326476484e-05,
+      "loss": 0.5851,
+      "step": 3225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2689404330600882,
+      "learning_rate": 1.7561757352155914e-05,
+      "loss": 0.6424,
+      "step": 3226
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2317719355357146,
+      "learning_rate": 1.7560112900422583e-05,
+      "loss": 0.6589,
+      "step": 3227
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3138338305062551,
+      "learning_rate": 1.7558467971380323e-05,
+      "loss": 0.5873,
+      "step": 3228
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1362952003679108,
+      "learning_rate": 1.755682256513298e-05,
+      "loss": 0.5645,
+      "step": 3229
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1628404898018452,
+      "learning_rate": 1.755517668178444e-05,
+      "loss": 0.6042,
+      "step": 3230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1885953168381038,
+      "learning_rate": 1.7553530321438625e-05,
+      "loss": 0.5527,
+      "step": 3231
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2439340273177062,
+      "learning_rate": 1.755188348419946e-05,
+      "loss": 0.6827,
+      "step": 3232
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3462039364413905,
+      "learning_rate": 1.755023617017093e-05,
+      "loss": 0.6536,
+      "step": 3233
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3549725946420983,
+      "learning_rate": 1.7548588379457042e-05,
+      "loss": 0.6188,
+      "step": 3234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3044959287422724,
+      "learning_rate": 1.754694011216182e-05,
+      "loss": 0.6692,
+      "step": 3235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.127275232465399,
+      "learning_rate": 1.754529136838933e-05,
+      "loss": 0.5306,
+      "step": 3236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1800491587049007,
+      "learning_rate": 1.7543642148243672e-05,
+      "loss": 0.6129,
+      "step": 3237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2309093352613774,
+      "learning_rate": 1.754199245182896e-05,
+      "loss": 0.623,
+      "step": 3238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1864438128343366,
+      "learning_rate": 1.7540342279249355e-05,
+      "loss": 0.5829,
+      "step": 3239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1907681116316102,
+      "learning_rate": 1.753869163060904e-05,
+      "loss": 0.6444,
+      "step": 3240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1033147308321871,
+      "learning_rate": 1.753704050601223e-05,
+      "loss": 0.6013,
+      "step": 3241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.119433033608672,
+      "learning_rate": 1.7535388905563163e-05,
+      "loss": 0.5642,
+      "step": 3242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1216088683289283,
+      "learning_rate": 1.7533736829366116e-05,
+      "loss": 0.5404,
+      "step": 3243
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3375103263812862,
+      "learning_rate": 1.7532084277525396e-05,
+      "loss": 0.6429,
+      "step": 3244
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2044718251208995,
+      "learning_rate": 1.7530431250145335e-05,
+      "loss": 0.63,
+      "step": 3245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.224135409051966,
+      "learning_rate": 1.7528777747330296e-05,
+      "loss": 0.6245,
+      "step": 3246
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1316154205917992,
+      "learning_rate": 1.7527123769184676e-05,
+      "loss": 0.5946,
+      "step": 3247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.238204713014442,
+      "learning_rate": 1.75254693158129e-05,
+      "loss": 0.5939,
+      "step": 3248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1776740468327798,
+      "learning_rate": 1.7523814387319413e-05,
+      "loss": 0.5823,
+      "step": 3249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1853791421449613,
+      "learning_rate": 1.752215898380871e-05,
+      "loss": 0.5543,
+      "step": 3250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3081901269439555,
+      "learning_rate": 1.7520503105385303e-05,
+      "loss": 0.6349,
+      "step": 3251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2068558846396564,
+      "learning_rate": 1.7518846752153732e-05,
+      "loss": 0.6398,
+      "step": 3252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1487598615519865,
+      "learning_rate": 1.7517189924218573e-05,
+      "loss": 0.5859,
+      "step": 3253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2408946765128526,
+      "learning_rate": 1.751553262168443e-05,
+      "loss": 0.6284,
+      "step": 3254
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.252214425323239,
+      "learning_rate": 1.751387484465594e-05,
+      "loss": 0.602,
+      "step": 3255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2074338371205104,
+      "learning_rate": 1.751221659323776e-05,
+      "loss": 0.592,
+      "step": 3256
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1190562058276843,
+      "learning_rate": 1.7510557867534594e-05,
+      "loss": 0.5727,
+      "step": 3257
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1997168186954494,
+      "learning_rate": 1.7508898667651158e-05,
+      "loss": 0.6496,
+      "step": 3258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.188626507944179,
+      "learning_rate": 1.7507238993692207e-05,
+      "loss": 0.6299,
+      "step": 3259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1822751698064238,
+      "learning_rate": 1.750557884576253e-05,
+      "loss": 0.6066,
+      "step": 3260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3028565768414537,
+      "learning_rate": 1.7503918223966932e-05,
+      "loss": 0.6039,
+      "step": 3261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1614333294057304,
+      "learning_rate": 1.7502257128410265e-05,
+      "loss": 0.6381,
+      "step": 3262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1393542069270943,
+      "learning_rate": 1.75005955591974e-05,
+      "loss": 0.5498,
+      "step": 3263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.231385340731515,
+      "learning_rate": 1.7498933516433233e-05,
+      "loss": 0.6238,
+      "step": 3264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1913341062621896,
+      "learning_rate": 1.749727100022271e-05,
+      "loss": 0.6228,
+      "step": 3265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2985829057310687,
+      "learning_rate": 1.7495608010670783e-05,
+      "loss": 0.684,
+      "step": 3266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1821546713263802,
+      "learning_rate": 1.7493944547882454e-05,
+      "loss": 0.5988,
+      "step": 3267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2065675208859925,
+      "learning_rate": 1.749228061196274e-05,
+      "loss": 0.5963,
+      "step": 3268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1694908303288254,
+      "learning_rate": 1.7490616203016696e-05,
+      "loss": 0.6042,
+      "step": 3269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.201746807671075,
+      "learning_rate": 1.7488951321149405e-05,
+      "loss": 0.6107,
+      "step": 3270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2412208298619856,
+      "learning_rate": 1.7487285966465982e-05,
+      "loss": 0.6231,
+      "step": 3271
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1691350822686961,
+      "learning_rate": 1.7485620139071564e-05,
+      "loss": 0.5284,
+      "step": 3272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1960277536674158,
+      "learning_rate": 1.7483953839071324e-05,
+      "loss": 0.614,
+      "step": 3273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.251885184647968,
+      "learning_rate": 1.748228706657047e-05,
+      "loss": 0.6663,
+      "step": 3274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1440939735182525,
+      "learning_rate": 1.7480619821674226e-05,
+      "loss": 0.5752,
+      "step": 3275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1080195783544828,
+      "learning_rate": 1.747895210448786e-05,
+      "loss": 0.5984,
+      "step": 3276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2818116143456204,
+      "learning_rate": 1.747728391511666e-05,
+      "loss": 0.6211,
+      "step": 3277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1357567245851952,
+      "learning_rate": 1.747561525366595e-05,
+      "loss": 0.5827,
+      "step": 3278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.277899245118622,
+      "learning_rate": 1.747394612024108e-05,
+      "loss": 0.6351,
+      "step": 3279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2762655204554059,
+      "learning_rate": 1.747227651494743e-05,
+      "loss": 0.618,
+      "step": 3280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2271707083784782,
+      "learning_rate": 1.747060643789041e-05,
+      "loss": 0.5917,
+      "step": 3281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.242839809901325,
+      "learning_rate": 1.7468935889175466e-05,
+      "loss": 0.5971,
+      "step": 3282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2811186188118213,
+      "learning_rate": 1.7467264868908064e-05,
+      "loss": 0.6199,
+      "step": 3283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1961908541480428,
+      "learning_rate": 1.7465593377193704e-05,
+      "loss": 0.6229,
+      "step": 3284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.355854518203587,
+      "learning_rate": 1.7463921414137916e-05,
+      "loss": 0.6819,
+      "step": 3285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2883341263097878,
+      "learning_rate": 1.746224897984626e-05,
+      "loss": 0.5962,
+      "step": 3286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2496667894658595,
+      "learning_rate": 1.7460576074424327e-05,
+      "loss": 0.6196,
+      "step": 3287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3655899741702564,
+      "learning_rate": 1.745890269797774e-05,
+      "loss": 0.6604,
+      "step": 3288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0832631504115473,
+      "learning_rate": 1.7457228850612132e-05,
+      "loss": 0.6096,
+      "step": 3289
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1328132366309567,
+      "learning_rate": 1.7455554532433198e-05,
+      "loss": 0.6114,
+      "step": 3290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.136917568791836,
+      "learning_rate": 1.745387974354664e-05,
+      "loss": 0.5363,
+      "step": 3291
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2180852789002845,
+      "learning_rate": 1.74522044840582e-05,
+      "loss": 0.5695,
+      "step": 3292
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.145202179678769,
+      "learning_rate": 1.7450528754073638e-05,
+      "loss": 0.5848,
+      "step": 3293
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2232268832993016,
+      "learning_rate": 1.744885255369876e-05,
+      "loss": 0.5984,
+      "step": 3294
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1450380108283569,
+      "learning_rate": 1.7447175883039386e-05,
+      "loss": 0.6363,
+      "step": 3295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3036661318962648,
+      "learning_rate": 1.744549874220138e-05,
+      "loss": 0.6107,
+      "step": 3296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2884760129336736,
+      "learning_rate": 1.744382113129062e-05,
+      "loss": 0.6404,
+      "step": 3297
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2740058800637986,
+      "learning_rate": 1.744214305041303e-05,
+      "loss": 0.6113,
+      "step": 3298
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1967480086642135,
+      "learning_rate": 1.7440464499674553e-05,
+      "loss": 0.6188,
+      "step": 3299
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3453225980550259,
+      "learning_rate": 1.7438785479181164e-05,
+      "loss": 0.6536,
+      "step": 3300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2878115045682026,
+      "learning_rate": 1.7437105989038868e-05,
+      "loss": 0.6355,
+      "step": 3301
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.171885833690157,
+      "learning_rate": 1.74354260293537e-05,
+      "loss": 0.5886,
+      "step": 3302
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4211772841721313,
+      "learning_rate": 1.7433745600231726e-05,
+      "loss": 0.6045,
+      "step": 3303
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3650233551794055,
+      "learning_rate": 1.743206470177904e-05,
+      "loss": 0.6654,
+      "step": 3304
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1674912921654297,
+      "learning_rate": 1.743038333410176e-05,
+      "loss": 0.6003,
+      "step": 3305
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2374112415272294,
+      "learning_rate": 1.7428701497306048e-05,
+      "loss": 0.6522,
+      "step": 3306
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2440135657377303,
+      "learning_rate": 1.7427019191498086e-05,
+      "loss": 0.576,
+      "step": 3307
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1372783109504145,
+      "learning_rate": 1.7425336416784082e-05,
+      "loss": 0.621,
+      "step": 3308
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.322163114419776,
+      "learning_rate": 1.7423653173270278e-05,
+      "loss": 0.6274,
+      "step": 3309
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.201498393517673,
+      "learning_rate": 1.742196946106295e-05,
+      "loss": 0.5946,
+      "step": 3310
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.160962835885136,
+      "learning_rate": 1.74202852802684e-05,
+      "loss": 0.594,
+      "step": 3311
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2198264430056878,
+      "learning_rate": 1.741860063099295e-05,
+      "loss": 0.5602,
+      "step": 3312
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.215235116603408,
+      "learning_rate": 1.7416915513342973e-05,
+      "loss": 0.5971,
+      "step": 3313
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2221801123204383,
+      "learning_rate": 1.7415229927424853e-05,
+      "loss": 0.6551,
+      "step": 3314
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.10844995320124,
+      "learning_rate": 1.741354387334501e-05,
+      "loss": 0.5566,
+      "step": 3315
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2094471281558543,
+      "learning_rate": 1.741185735120989e-05,
+      "loss": 0.6702,
+      "step": 3316
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2872892725812624,
+      "learning_rate": 1.7410170361125978e-05,
+      "loss": 0.6408,
+      "step": 3317
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.125605685054057,
+      "learning_rate": 1.740848290319978e-05,
+      "loss": 0.5682,
+      "step": 3318
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3357768268893488,
+      "learning_rate": 1.7406794977537832e-05,
+      "loss": 0.6534,
+      "step": 3319
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2486897754370647,
+      "learning_rate": 1.7405106584246705e-05,
+      "loss": 0.6187,
+      "step": 3320
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1170026186038942,
+      "learning_rate": 1.740341772343299e-05,
+      "loss": 0.5665,
+      "step": 3321
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1826839664420352,
+      "learning_rate": 1.7401728395203323e-05,
+      "loss": 0.5946,
+      "step": 3322
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.081077954490754,
+      "learning_rate": 1.7400038599664354e-05,
+      "loss": 0.5587,
+      "step": 3323
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3094195137560254,
+      "learning_rate": 1.7398348336922764e-05,
+      "loss": 0.6296,
+      "step": 3324
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2678910678799522,
+      "learning_rate": 1.7396657607085276e-05,
+      "loss": 0.6614,
+      "step": 3325
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2309330624308885,
+      "learning_rate": 1.739496641025863e-05,
+      "loss": 0.6511,
+      "step": 3326
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1810842670252255,
+      "learning_rate": 1.7393274746549605e-05,
+      "loss": 0.6624,
+      "step": 3327
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0927437649568406,
+      "learning_rate": 1.7391582616064998e-05,
+      "loss": 0.5657,
+      "step": 3328
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1824833158642718,
+      "learning_rate": 1.7389890018911647e-05,
+      "loss": 0.6251,
+      "step": 3329
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2882377066578126,
+      "learning_rate": 1.738819695519641e-05,
+      "loss": 0.6431,
+      "step": 3330
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.274624464417439,
+      "learning_rate": 1.7386503425026183e-05,
+      "loss": 0.6202,
+      "step": 3331
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.204040278277466,
+      "learning_rate": 1.7384809428507884e-05,
+      "loss": 0.5557,
+      "step": 3332
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.197070854237926,
+      "learning_rate": 1.7383114965748465e-05,
+      "loss": 0.6166,
+      "step": 3333
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3553243774569124,
+      "learning_rate": 1.738142003685491e-05,
+      "loss": 0.7191,
+      "step": 3334
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1880538552308022,
+      "learning_rate": 1.737972464193422e-05,
+      "loss": 0.5912,
+      "step": 3335
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2519953061524662,
+      "learning_rate": 1.7378028781093443e-05,
+      "loss": 0.6046,
+      "step": 3336
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2172224301772756,
+      "learning_rate": 1.7376332454439643e-05,
+      "loss": 0.555,
+      "step": 3337
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2924892769669403,
+      "learning_rate": 1.7374635662079915e-05,
+      "loss": 0.6675,
+      "step": 3338
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3047546095349118,
+      "learning_rate": 1.7372938404121393e-05,
+      "loss": 0.6306,
+      "step": 3339
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3009045058498545,
+      "learning_rate": 1.7371240680671228e-05,
+      "loss": 0.6191,
+      "step": 3340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2138590854610696,
+      "learning_rate": 1.7369542491836608e-05,
+      "loss": 0.5735,
+      "step": 3341
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2044767242374326,
+      "learning_rate": 1.736784383772475e-05,
+      "loss": 0.6026,
+      "step": 3342
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1745416944596863,
+      "learning_rate": 1.7366144718442893e-05,
+      "loss": 0.6085,
+      "step": 3343
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2627075855576206,
+      "learning_rate": 1.736444513409832e-05,
+      "loss": 0.6771,
+      "step": 3344
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1491306833927326,
+      "learning_rate": 1.736274508479833e-05,
+      "loss": 0.5387,
+      "step": 3345
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1292567253562806,
+      "learning_rate": 1.7361044570650256e-05,
+      "loss": 0.557,
+      "step": 3346
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3335490151641733,
+      "learning_rate": 1.735934359176146e-05,
+      "loss": 0.5979,
+      "step": 3347
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1763961552817495,
+      "learning_rate": 1.7357642148239334e-05,
+      "loss": 0.5747,
+      "step": 3348
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1725573778245444,
+      "learning_rate": 1.73559402401913e-05,
+      "loss": 0.5861,
+      "step": 3349
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1778224829870367,
+      "learning_rate": 1.7354237867724805e-05,
+      "loss": 0.5585,
+      "step": 3350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1484453888706514,
+      "learning_rate": 1.7352535030947334e-05,
+      "loss": 0.5919,
+      "step": 3351
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2300903707350819,
+      "learning_rate": 1.735083172996639e-05,
+      "loss": 0.6022,
+      "step": 3352
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2197658022454867,
+      "learning_rate": 1.7349127964889508e-05,
+      "loss": 0.6532,
+      "step": 3353
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.231761676950544,
+      "learning_rate": 1.7347423735824266e-05,
+      "loss": 0.6267,
+      "step": 3354
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2859242008675675,
+      "learning_rate": 1.734571904287826e-05,
+      "loss": 0.6163,
+      "step": 3355
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.178086514438827,
+      "learning_rate": 1.7344013886159104e-05,
+      "loss": 0.617,
+      "step": 3356
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3033751323140503,
+      "learning_rate": 1.7342308265774467e-05,
+      "loss": 0.6533,
+      "step": 3357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2567069366100936,
+      "learning_rate": 1.7340602181832028e-05,
+      "loss": 0.6165,
+      "step": 3358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2007666920143658,
+      "learning_rate": 1.7338895634439496e-05,
+      "loss": 0.6301,
+      "step": 3359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2706204480991743,
+      "learning_rate": 1.733718862370462e-05,
+      "loss": 0.6285,
+      "step": 3360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3147399498714427,
+      "learning_rate": 1.7335481149735173e-05,
+      "loss": 0.6519,
+      "step": 3361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.228480934417426,
+      "learning_rate": 1.7333773212638957e-05,
+      "loss": 0.5986,
+      "step": 3362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1688612777855818,
+      "learning_rate": 1.73320648125238e-05,
+      "loss": 0.5989,
+      "step": 3363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1706408296035187,
+      "learning_rate": 1.733035594949756e-05,
+      "loss": 0.5714,
+      "step": 3364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.184445367499233,
+      "learning_rate": 1.732864662366813e-05,
+      "loss": 0.5902,
+      "step": 3365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2335445652748194,
+      "learning_rate": 1.7326936835143427e-05,
+      "loss": 0.5393,
+      "step": 3366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2273502128717442,
+      "learning_rate": 1.73252265840314e-05,
+      "loss": 0.6452,
+      "step": 3367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2024560034429803,
+      "learning_rate": 1.7323515870440027e-05,
+      "loss": 0.626,
+      "step": 3368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.094234086947422,
+      "learning_rate": 1.7321804694477314e-05,
+      "loss": 0.5842,
+      "step": 3369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2529117050723875,
+      "learning_rate": 1.7320093056251293e-05,
+      "loss": 0.6102,
+      "step": 3370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1553890528349962,
+      "learning_rate": 1.7318380955870032e-05,
+      "loss": 0.5824,
+      "step": 3371
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1919226856024154,
+      "learning_rate": 1.7316668393441622e-05,
+      "loss": 0.6054,
+      "step": 3372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3072211967208809,
+      "learning_rate": 1.731495536907419e-05,
+      "loss": 0.6519,
+      "step": 3373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3214918207990012,
+      "learning_rate": 1.7313241882875883e-05,
+      "loss": 0.6443,
+      "step": 3374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.15111137052502,
+      "learning_rate": 1.7311527934954885e-05,
+      "loss": 0.5764,
+      "step": 3375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2135490067987205,
+      "learning_rate": 1.7309813525419403e-05,
+      "loss": 0.6,
+      "step": 3376
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0840930781116456,
+      "learning_rate": 1.730809865437768e-05,
+      "loss": 0.5167,
+      "step": 3377
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2688477069834467,
+      "learning_rate": 1.7306383321937986e-05,
+      "loss": 0.6351,
+      "step": 3378
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.141371495777199,
+      "learning_rate": 1.730466752820862e-05,
+      "loss": 0.6126,
+      "step": 3379
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.304618468142981,
+      "learning_rate": 1.7302951273297904e-05,
+      "loss": 0.6449,
+      "step": 3380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1519705992596048,
+      "learning_rate": 1.7301234557314194e-05,
+      "loss": 0.5883,
+      "step": 3381
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.200634248331523,
+      "learning_rate": 1.7299517380365877e-05,
+      "loss": 0.6639,
+      "step": 3382
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2580156339968729,
+      "learning_rate": 1.7297799742561367e-05,
+      "loss": 0.5991,
+      "step": 3383
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2387512951487094,
+      "learning_rate": 1.729608164400911e-05,
+      "loss": 0.5469,
+      "step": 3384
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2410634551228308,
+      "learning_rate": 1.7294363084817573e-05,
+      "loss": 0.6314,
+      "step": 3385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2780469539476047,
+      "learning_rate": 1.7292644065095263e-05,
+      "loss": 0.6622,
+      "step": 3386
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2191027595386779,
+      "learning_rate": 1.7290924584950704e-05,
+      "loss": 0.5835,
+      "step": 3387
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1113196508423067,
+      "learning_rate": 1.7289204644492463e-05,
+      "loss": 0.6007,
+      "step": 3388
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1969095668332939,
+      "learning_rate": 1.7287484243829126e-05,
+      "loss": 0.6494,
+      "step": 3389
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1119038945732729,
+      "learning_rate": 1.728576338306931e-05,
+      "loss": 0.6113,
+      "step": 3390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.302236069939979,
+      "learning_rate": 1.7284042062321663e-05,
+      "loss": 0.6146,
+      "step": 3391
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1797731886304792,
+      "learning_rate": 1.7282320281694857e-05,
+      "loss": 0.5641,
+      "step": 3392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.254508851136797,
+      "learning_rate": 1.72805980412976e-05,
+      "loss": 0.597,
+      "step": 3393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3392448310061567,
+      "learning_rate": 1.7278875341238627e-05,
+      "loss": 0.7139,
+      "step": 3394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3181460059541965,
+      "learning_rate": 1.7277152181626703e-05,
+      "loss": 0.6158,
+      "step": 3395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2433147473944648,
+      "learning_rate": 1.727542856257061e-05,
+      "loss": 0.6183,
+      "step": 3396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2530361496017832,
+      "learning_rate": 1.727370448417918e-05,
+      "loss": 0.6041,
+      "step": 3397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.244108094848589,
+      "learning_rate": 1.7271979946561256e-05,
+      "loss": 0.611,
+      "step": 3398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2718321153313898,
+      "learning_rate": 1.7270254949825722e-05,
+      "loss": 0.602,
+      "step": 3399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1197390462650894,
+      "learning_rate": 1.726852949408148e-05,
+      "loss": 0.548,
+      "step": 3400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1633571561562022,
+      "learning_rate": 1.7266803579437472e-05,
+      "loss": 0.6289,
+      "step": 3401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2020785747590357,
+      "learning_rate": 1.7265077206002664e-05,
+      "loss": 0.6174,
+      "step": 3402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1912704640539116,
+      "learning_rate": 1.7263350373886046e-05,
+      "loss": 0.5959,
+      "step": 3403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.212634810592529,
+      "learning_rate": 1.726162308319665e-05,
+      "loss": 0.6457,
+      "step": 3404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1948400822978884,
+      "learning_rate": 1.7259895334043516e-05,
+      "loss": 0.6085,
+      "step": 3405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.269153010436185,
+      "learning_rate": 1.725816712653574e-05,
+      "loss": 0.6412,
+      "step": 3406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.176745402796988,
+      "learning_rate": 1.7256438460782427e-05,
+      "loss": 0.6127,
+      "step": 3407
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1200628026316155,
+      "learning_rate": 1.725470933689271e-05,
+      "loss": 0.488,
+      "step": 3408
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2441838854300658,
+      "learning_rate": 1.7252979754975765e-05,
+      "loss": 0.6186,
+      "step": 3409
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.190531375037393,
+      "learning_rate": 1.725124971514079e-05,
+      "loss": 0.6223,
+      "step": 3410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.231457122576695,
+      "learning_rate": 1.7249519217497007e-05,
+      "loss": 0.6121,
+      "step": 3411
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1829266572436101,
+      "learning_rate": 1.7247788262153673e-05,
+      "loss": 0.6121,
+      "step": 3412
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.290336949899096,
+      "learning_rate": 1.724605684922007e-05,
+      "loss": 0.6457,
+      "step": 3413
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.253085714661441,
+      "learning_rate": 1.7244324978805516e-05,
+      "loss": 0.6617,
+      "step": 3414
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4160945054807461,
+      "learning_rate": 1.7242592651019353e-05,
+      "loss": 0.6496,
+      "step": 3415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1085313624865156,
+      "learning_rate": 1.7240859865970948e-05,
+      "loss": 0.5781,
+      "step": 3416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1939958289311587,
+      "learning_rate": 1.7239126623769703e-05,
+      "loss": 0.6262,
+      "step": 3417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2282199716125397,
+      "learning_rate": 1.7237392924525037e-05,
+      "loss": 0.6242,
+      "step": 3418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2968295905194807,
+      "learning_rate": 1.7235658768346422e-05,
+      "loss": 0.6765,
+      "step": 3419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2039937932529343,
+      "learning_rate": 1.723392415534334e-05,
+      "loss": 0.551,
+      "step": 3420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.182970392732156,
+      "learning_rate": 1.72321890856253e-05,
+      "loss": 0.5846,
+      "step": 3421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2279794368917152,
+      "learning_rate": 1.723045355930185e-05,
+      "loss": 0.5978,
+      "step": 3422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1248688621383214,
+      "learning_rate": 1.7228717576482563e-05,
+      "loss": 0.5783,
+      "step": 3423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3525211704874793,
+      "learning_rate": 1.722698113727704e-05,
+      "loss": 0.6685,
+      "step": 3424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1919329370220593,
+      "learning_rate": 1.7225244241794916e-05,
+      "loss": 0.528,
+      "step": 3425
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2538218722370853,
+      "learning_rate": 1.7223506890145842e-05,
+      "loss": 0.6043,
+      "step": 3426
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1875692648513518,
+      "learning_rate": 1.7221769082439508e-05,
+      "loss": 0.5975,
+      "step": 3427
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2182915143056035,
+      "learning_rate": 1.7220030818785635e-05,
+      "loss": 0.6354,
+      "step": 3428
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2227457519532332,
+      "learning_rate": 1.721829209929396e-05,
+      "loss": 0.6272,
+      "step": 3429
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.271587361795596,
+      "learning_rate": 1.721655292407427e-05,
+      "loss": 0.684,
+      "step": 3430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1728294046486247,
+      "learning_rate": 1.721481329323636e-05,
+      "loss": 0.5759,
+      "step": 3431
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2359590627378376,
+      "learning_rate": 1.7213073206890063e-05,
+      "loss": 0.6133,
+      "step": 3432
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2561826868195167,
+      "learning_rate": 1.721133266514524e-05,
+      "loss": 0.6566,
+      "step": 3433
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1822220309864055,
+      "learning_rate": 1.720959166811178e-05,
+      "loss": 0.6348,
+      "step": 3434
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3352234476945937,
+      "learning_rate": 1.72078502158996e-05,
+      "loss": 0.6142,
+      "step": 3435
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2966064266947603,
+      "learning_rate": 1.720610830861865e-05,
+      "loss": 0.6492,
+      "step": 3436
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0881068564386256,
+      "learning_rate": 1.7204365946378906e-05,
+      "loss": 0.5589,
+      "step": 3437
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.13118300476648,
+      "learning_rate": 1.7202623129290367e-05,
+      "loss": 0.5947,
+      "step": 3438
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2141565176007922,
+      "learning_rate": 1.720087985746307e-05,
+      "loss": 0.5997,
+      "step": 3439
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2549494506716794,
+      "learning_rate": 1.7199136131007074e-05,
+      "loss": 0.5939,
+      "step": 3440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1754788234755198,
+      "learning_rate": 1.719739195003247e-05,
+      "loss": 0.6065,
+      "step": 3441
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1796197934962993,
+      "learning_rate": 1.7195647314649383e-05,
+      "loss": 0.599,
+      "step": 3442
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2069171740503573,
+      "learning_rate": 1.7193902224967956e-05,
+      "loss": 0.6181,
+      "step": 3443
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3393779868174052,
+      "learning_rate": 1.7192156681098364e-05,
+      "loss": 0.5985,
+      "step": 3444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1614857771614773,
+      "learning_rate": 1.7190410683150816e-05,
+      "loss": 0.5979,
+      "step": 3445
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.221602207962269,
+      "learning_rate": 1.7188664231235544e-05,
+      "loss": 0.5734,
+      "step": 3446
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2435657844030366,
+      "learning_rate": 1.7186917325462808e-05,
+      "loss": 0.6279,
+      "step": 3447
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.252003209009415,
+      "learning_rate": 1.71851699659429e-05,
+      "loss": 0.6782,
+      "step": 3448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.168796054932257,
+      "learning_rate": 1.7183422152786145e-05,
+      "loss": 0.5885,
+      "step": 3449
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1936584697205102,
+      "learning_rate": 1.718167388610289e-05,
+      "loss": 0.5842,
+      "step": 3450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1738174107930588,
+      "learning_rate": 1.7179925166003506e-05,
+      "loss": 0.5997,
+      "step": 3451
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1548514284630382,
+      "learning_rate": 1.71781759925984e-05,
+      "loss": 0.5791,
+      "step": 3452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.23274374263663,
+      "learning_rate": 1.7176426365998015e-05,
+      "loss": 0.6286,
+      "step": 3453
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1556862668290826,
+      "learning_rate": 1.7174676286312807e-05,
+      "loss": 0.5293,
+      "step": 3454
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2823091641597584,
+      "learning_rate": 1.717292575365327e-05,
+      "loss": 0.6569,
+      "step": 3455
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2788120428492529,
+      "learning_rate": 1.717117476812992e-05,
+      "loss": 0.6217,
+      "step": 3456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3612913020009727,
+      "learning_rate": 1.7169423329853307e-05,
+      "loss": 0.687,
+      "step": 3457
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3048652482160341,
+      "learning_rate": 1.7167671438934014e-05,
+      "loss": 0.5577,
+      "step": 3458
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.127044674015843,
+      "learning_rate": 1.7165919095482636e-05,
+      "loss": 0.5872,
+      "step": 3459
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1676144780267974,
+      "learning_rate": 1.716416629960982e-05,
+      "loss": 0.6488,
+      "step": 3460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.365012744382953,
+      "learning_rate": 1.7162413051426222e-05,
+      "loss": 0.6613,
+      "step": 3461
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0879256895156963,
+      "learning_rate": 1.7160659351042533e-05,
+      "loss": 0.5603,
+      "step": 3462
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1720538193645633,
+      "learning_rate": 1.7158905198569476e-05,
+      "loss": 0.5483,
+      "step": 3463
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.23895635162207,
+      "learning_rate": 1.7157150594117805e-05,
+      "loss": 0.6116,
+      "step": 3464
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1681729651696895,
+      "learning_rate": 1.7155395537798282e-05,
+      "loss": 0.5921,
+      "step": 3465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2154166781050915,
+      "learning_rate": 1.7153640029721726e-05,
+      "loss": 0.6222,
+      "step": 3466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2614323908151381,
+      "learning_rate": 1.7151884069998966e-05,
+      "loss": 0.6172,
+      "step": 3467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1960065235546904,
+      "learning_rate": 1.7150127658740868e-05,
+      "loss": 0.5752,
+      "step": 3468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2162324383086238,
+      "learning_rate": 1.7148370796058316e-05,
+      "loss": 0.6484,
+      "step": 3469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0929767054826567,
+      "learning_rate": 1.714661348206224e-05,
+      "loss": 0.5337,
+      "step": 3470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2129818303067432,
+      "learning_rate": 1.714485571686358e-05,
+      "loss": 0.6408,
+      "step": 3471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2663385710849338,
+      "learning_rate": 1.7143097500573314e-05,
+      "loss": 0.6163,
+      "step": 3472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3020229427320296,
+      "learning_rate": 1.7141338833302454e-05,
+      "loss": 0.6458,
+      "step": 3473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2019619955951104,
+      "learning_rate": 1.713957971516203e-05,
+      "loss": 0.6255,
+      "step": 3474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2143210117435042,
+      "learning_rate": 1.71378201462631e-05,
+      "loss": 0.5798,
+      "step": 3475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2721459795989332,
+      "learning_rate": 1.7136060126716756e-05,
+      "loss": 0.6014,
+      "step": 3476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1761004752814077,
+      "learning_rate": 1.7134299656634124e-05,
+      "loss": 0.5792,
+      "step": 3477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1795035717802151,
+      "learning_rate": 1.7132538736126342e-05,
+      "loss": 0.6121,
+      "step": 3478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2644927529202392,
+      "learning_rate": 1.713077736530459e-05,
+      "loss": 0.6117,
+      "step": 3479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2594005913417883,
+      "learning_rate": 1.712901554428008e-05,
+      "loss": 0.6772,
+      "step": 3480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3499656902474506,
+      "learning_rate": 1.7127253273164032e-05,
+      "loss": 0.6002,
+      "step": 3481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2425175834040978,
+      "learning_rate": 1.7125490552067713e-05,
+      "loss": 0.6348,
+      "step": 3482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1711106414289294,
+      "learning_rate": 1.7123727381102417e-05,
+      "loss": 0.6374,
+      "step": 3483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1554935145738767,
+      "learning_rate": 1.7121963760379453e-05,
+      "loss": 0.5657,
+      "step": 3484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6971287995949367,
+      "learning_rate": 1.7120199690010176e-05,
+      "loss": 0.6261,
+      "step": 3485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1980711077219164,
+      "learning_rate": 1.7118435170105955e-05,
+      "loss": 0.6241,
+      "step": 3486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2369071484397642,
+      "learning_rate": 1.7116670200778192e-05,
+      "loss": 0.6096,
+      "step": 3487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.143335805219271,
+      "learning_rate": 1.711490478213833e-05,
+      "loss": 0.5794,
+      "step": 3488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3044845057656975,
+      "learning_rate": 1.7113138914297817e-05,
+      "loss": 0.5988,
+      "step": 3489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2158888464306223,
+      "learning_rate": 1.7111372597368143e-05,
+      "loss": 0.6161,
+      "step": 3490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1621666132831512,
+      "learning_rate": 1.7109605831460833e-05,
+      "loss": 0.597,
+      "step": 3491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2148988472075475,
+      "learning_rate": 1.710783861668742e-05,
+      "loss": 0.5849,
+      "step": 3492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2343771246396866,
+      "learning_rate": 1.7106070953159487e-05,
+      "loss": 0.6387,
+      "step": 3493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1944441789188138,
+      "learning_rate": 1.710430284098863e-05,
+      "loss": 0.5823,
+      "step": 3494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1677784844503332,
+      "learning_rate": 1.7102534280286483e-05,
+      "loss": 0.5873,
+      "step": 3495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1186957149015377,
+      "learning_rate": 1.71007652711647e-05,
+      "loss": 0.5679,
+      "step": 3496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2166223293656484,
+      "learning_rate": 1.7098995813734974e-05,
+      "loss": 0.6354,
+      "step": 3497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2219242915286959,
+      "learning_rate": 1.7097225908109015e-05,
+      "loss": 0.6316,
+      "step": 3498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1823154508688276,
+      "learning_rate": 1.7095455554398564e-05,
+      "loss": 0.5865,
+      "step": 3499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2132324130318173,
+      "learning_rate": 1.70936847527154e-05,
+      "loss": 0.5596,
+      "step": 3500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2081526697959808,
+      "learning_rate": 1.709191350317132e-05,
+      "loss": 0.6299,
+      "step": 3501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1080447535927218,
+      "learning_rate": 1.709014180587815e-05,
+      "loss": 0.5576,
+      "step": 3502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2654575189991,
+      "learning_rate": 1.7088369660947743e-05,
+      "loss": 0.6246,
+      "step": 3503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1062880536447204,
+      "learning_rate": 1.7086597068491994e-05,
+      "loss": 0.5676,
+      "step": 3504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1743979697343756,
+      "learning_rate": 1.7084824028622807e-05,
+      "loss": 0.6279,
+      "step": 3505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2516906272653687,
+      "learning_rate": 1.708305054145213e-05,
+      "loss": 0.6408,
+      "step": 3506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.112168194012932,
+      "learning_rate": 1.7081276607091925e-05,
+      "loss": 0.5922,
+      "step": 3507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2045648059631897,
+      "learning_rate": 1.7079502225654192e-05,
+      "loss": 0.5833,
+      "step": 3508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2075480123416586,
+      "learning_rate": 1.707772739725096e-05,
+      "loss": 0.525,
+      "step": 3509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.178969006545972,
+      "learning_rate": 1.7075952121994282e-05,
+      "loss": 0.5985,
+      "step": 3510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2524471171287925,
+      "learning_rate": 1.707417639999624e-05,
+      "loss": 0.6069,
+      "step": 3511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2340928250120227,
+      "learning_rate": 1.707240023136894e-05,
+      "loss": 0.5795,
+      "step": 3512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1014666684718406,
+      "learning_rate": 1.7070623616224528e-05,
+      "loss": 0.578,
+      "step": 3513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1604010840601888,
+      "learning_rate": 1.7068846554675166e-05,
+      "loss": 0.5926,
+      "step": 3514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.293871745771315,
+      "learning_rate": 1.706706904683305e-05,
+      "loss": 0.6135,
+      "step": 3515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1432762165279406,
+      "learning_rate": 1.7065291092810406e-05,
+      "loss": 0.5393,
+      "step": 3516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.262780890990764,
+      "learning_rate": 1.7063512692719482e-05,
+      "loss": 0.6187,
+      "step": 3517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1681728120982255,
+      "learning_rate": 1.7061733846672562e-05,
+      "loss": 0.5414,
+      "step": 3518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1907958921806925,
+      "learning_rate": 1.7059954554781945e-05,
+      "loss": 0.5693,
+      "step": 3519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2319715254391224,
+      "learning_rate": 1.7058174817159973e-05,
+      "loss": 0.6132,
+      "step": 3520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.240147577086982,
+      "learning_rate": 1.7056394633919012e-05,
+      "loss": 0.6046,
+      "step": 3521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5087442632153663,
+      "learning_rate": 1.705461400517145e-05,
+      "loss": 0.5853,
+      "step": 3522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4098506184169337,
+      "learning_rate": 1.705283293102971e-05,
+      "loss": 0.5993,
+      "step": 3523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2521401680706161,
+      "learning_rate": 1.7051051411606238e-05,
+      "loss": 0.5966,
+      "step": 3524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2914691076656413,
+      "learning_rate": 1.7049269447013515e-05,
+      "loss": 0.6186,
+      "step": 3525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3109645945280441,
+      "learning_rate": 1.704748703736404e-05,
+      "loss": 0.6218,
+      "step": 3526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2406136000563492,
+      "learning_rate": 1.7045704182770346e-05,
+      "loss": 0.5841,
+      "step": 3527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.267768974807635,
+      "learning_rate": 1.7043920883344998e-05,
+      "loss": 0.5947,
+      "step": 3528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2891440163471761,
+      "learning_rate": 1.7042137139200583e-05,
+      "loss": 0.6555,
+      "step": 3529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2079516608248553,
+      "learning_rate": 1.7040352950449716e-05,
+      "loss": 0.6532,
+      "step": 3530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1711107432205794,
+      "learning_rate": 1.7038568317205045e-05,
+      "loss": 0.5984,
+      "step": 3531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2320295335059828,
+      "learning_rate": 1.7036783239579243e-05,
+      "loss": 0.6114,
+      "step": 3532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3518348816285979,
+      "learning_rate": 1.703499771768501e-05,
+      "loss": 0.6722,
+      "step": 3533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.419602474930434,
+      "learning_rate": 1.7033211751635074e-05,
+      "loss": 0.6368,
+      "step": 3534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.15933993999819,
+      "learning_rate": 1.7031425341542193e-05,
+      "loss": 0.6554,
+      "step": 3535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2699663561021641,
+      "learning_rate": 1.7029638487519155e-05,
+      "loss": 0.6151,
+      "step": 3536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3049581098592968,
+      "learning_rate": 1.702785118967877e-05,
+      "loss": 0.622,
+      "step": 3537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2158346274885192,
+      "learning_rate": 1.702606344813388e-05,
+      "loss": 0.5997,
+      "step": 3538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1389321083820854,
+      "learning_rate": 1.7024275262997358e-05,
+      "loss": 0.5108,
+      "step": 3539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6805761780887818,
+      "learning_rate": 1.702248663438209e-05,
+      "loss": 0.6208,
+      "step": 3540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3455553125886481,
+      "learning_rate": 1.7020697562401017e-05,
+      "loss": 0.5675,
+      "step": 3541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0953874592292865,
+      "learning_rate": 1.7018908047167083e-05,
+      "loss": 0.5658,
+      "step": 3542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1857180526808158,
+      "learning_rate": 1.7017118088793267e-05,
+      "loss": 0.609,
+      "step": 3543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.150018343571857,
+      "learning_rate": 1.701532768739259e-05,
+      "loss": 0.5614,
+      "step": 3544
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2616653667293758,
+      "learning_rate": 1.7013536843078077e-05,
+      "loss": 0.6334,
+      "step": 3545
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2122241168430332,
+      "learning_rate": 1.70117455559628e-05,
+      "loss": 0.5925,
+      "step": 3546
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.171793820430402,
+      "learning_rate": 1.700995382615985e-05,
+      "loss": 0.5664,
+      "step": 3547
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1670900439489724,
+      "learning_rate": 1.7008161653782344e-05,
+      "loss": 0.5695,
+      "step": 3548
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1869272055250546,
+      "learning_rate": 1.7006369038943443e-05,
+      "loss": 0.5299,
+      "step": 3549
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1249377975221864,
+      "learning_rate": 1.700457598175631e-05,
+      "loss": 0.5286,
+      "step": 3550
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2016604042595487,
+      "learning_rate": 1.700278248233416e-05,
+      "loss": 0.6292,
+      "step": 3551
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1327456224365782,
+      "learning_rate": 1.700098854079022e-05,
+      "loss": 0.5653,
+      "step": 3552
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.146229860301497,
+      "learning_rate": 1.6999194157237753e-05,
+      "loss": 0.5406,
+      "step": 3553
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2107167196510564,
+      "learning_rate": 1.699739933179005e-05,
+      "loss": 0.6163,
+      "step": 3554
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.128816013623664,
+      "learning_rate": 1.6995604064560426e-05,
+      "loss": 0.5878,
+      "step": 3555
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2362328563461915,
+      "learning_rate": 1.699380835566222e-05,
+      "loss": 0.567,
+      "step": 3556
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1748535450505353,
+      "learning_rate": 1.6992012205208814e-05,
+      "loss": 0.6064,
+      "step": 3557
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2243381911815883,
+      "learning_rate": 1.6990215613313602e-05,
+      "loss": 0.5744,
+      "step": 3558
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2039026001634172,
+      "learning_rate": 1.6988418580090013e-05,
+      "loss": 0.6304,
+      "step": 3559
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1472622313821053,
+      "learning_rate": 1.6986621105651505e-05,
+      "loss": 0.5126,
+      "step": 3560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2402514358908967,
+      "learning_rate": 1.6984823190111558e-05,
+      "loss": 0.6169,
+      "step": 3561
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2778025978322296,
+      "learning_rate": 1.698302483358369e-05,
+      "loss": 0.5921,
+      "step": 3562
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1746282659089236,
+      "learning_rate": 1.6981226036181433e-05,
+      "loss": 0.6314,
+      "step": 3563
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2409812300505223,
+      "learning_rate": 1.6979426798018356e-05,
+      "loss": 0.5945,
+      "step": 3564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.140078322162819,
+      "learning_rate": 1.697762711920806e-05,
+      "loss": 0.5977,
+      "step": 3565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.327162854114233,
+      "learning_rate": 1.697582699986416e-05,
+      "loss": 0.6625,
+      "step": 3566
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.195544500836556,
+      "learning_rate": 1.697402644010032e-05,
+      "loss": 0.59,
+      "step": 3567
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2330646561611454,
+      "learning_rate": 1.6972225440030203e-05,
+      "loss": 0.5952,
+      "step": 3568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3666670326294448,
+      "learning_rate": 1.697042399976752e-05,
+      "loss": 0.6384,
+      "step": 3569
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1439809680636073,
+      "learning_rate": 1.6968622119426013e-05,
+      "loss": 0.5658,
+      "step": 3570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.349977611426743,
+      "learning_rate": 1.696681979911943e-05,
+      "loss": 0.6397,
+      "step": 3571
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1624509165256394,
+      "learning_rate": 1.696501703896158e-05,
+      "loss": 0.6571,
+      "step": 3572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1602184651451246,
+      "learning_rate": 1.6963213839066263e-05,
+      "loss": 0.5956,
+      "step": 3573
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1053391774942984,
+      "learning_rate": 1.696141019954733e-05,
+      "loss": 0.5249,
+      "step": 3574
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.093012588418014,
+      "learning_rate": 1.6959606120518656e-05,
+      "loss": 0.5847,
+      "step": 3575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.136895077587339,
+      "learning_rate": 1.695780160209414e-05,
+      "loss": 0.5928,
+      "step": 3576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.148027878974376,
+      "learning_rate": 1.6955996644387715e-05,
+      "loss": 0.6394,
+      "step": 3577
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2532928963150862,
+      "learning_rate": 1.695419124751333e-05,
+      "loss": 0.6068,
+      "step": 3578
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1299222710131411,
+      "learning_rate": 1.6952385411584974e-05,
+      "loss": 0.6031,
+      "step": 3579
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.193586362238359,
+      "learning_rate": 1.695057913671666e-05,
+      "loss": 0.5865,
+      "step": 3580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2034823890066315,
+      "learning_rate": 1.694877242302242e-05,
+      "loss": 0.5917,
+      "step": 3581
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2284215457475098,
+      "learning_rate": 1.694696527061633e-05,
+      "loss": 0.6301,
+      "step": 3582
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1804113630361248,
+      "learning_rate": 1.6945157679612478e-05,
+      "loss": 0.5869,
+      "step": 3583
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.251924511467942,
+      "learning_rate": 1.694334965012499e-05,
+      "loss": 0.6105,
+      "step": 3584
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2661377139075032,
+      "learning_rate": 1.6941541182268015e-05,
+      "loss": 0.5938,
+      "step": 3585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2739834697465422,
+      "learning_rate": 1.6939732276155733e-05,
+      "loss": 0.5946,
+      "step": 3586
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1314964278936366,
+      "learning_rate": 1.6937922931902348e-05,
+      "loss": 0.5699,
+      "step": 3587
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1397968578040283,
+      "learning_rate": 1.6936113149622093e-05,
+      "loss": 0.5665,
+      "step": 3588
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1232114986339181,
+      "learning_rate": 1.6934302929429226e-05,
+      "loss": 0.5357,
+      "step": 3589
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2606855478720311,
+      "learning_rate": 1.6932492271438046e-05,
+      "loss": 0.6386,
+      "step": 3590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2709180094957568,
+      "learning_rate": 1.6930681175762855e-05,
+      "loss": 0.6314,
+      "step": 3591
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1488307876322585,
+      "learning_rate": 1.692886964251801e-05,
+      "loss": 0.5412,
+      "step": 3592
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1900822020436885,
+      "learning_rate": 1.6927057671817872e-05,
+      "loss": 0.5871,
+      "step": 3593
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1997646120790852,
+      "learning_rate": 1.6925245263776842e-05,
+      "loss": 0.6134,
+      "step": 3594
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.15087450404533,
+      "learning_rate": 1.6923432418509356e-05,
+      "loss": 0.6305,
+      "step": 3595
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.150878543716134,
+      "learning_rate": 1.6921619136129856e-05,
+      "loss": 0.6098,
+      "step": 3596
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2434921131474457,
+      "learning_rate": 1.691980541675283e-05,
+      "loss": 0.6404,
+      "step": 3597
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3097038456982897,
+      "learning_rate": 1.6917991260492787e-05,
+      "loss": 0.611,
+      "step": 3598
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.229576829052618,
+      "learning_rate": 1.691617666746426e-05,
+      "loss": 0.5951,
+      "step": 3599
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2509677478679255,
+      "learning_rate": 1.691436163778182e-05,
+      "loss": 0.6254,
+      "step": 3600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2207993126166072,
+      "learning_rate": 1.691254617156006e-05,
+      "loss": 0.6067,
+      "step": 3601
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2126412496190742,
+      "learning_rate": 1.6910730268913593e-05,
+      "loss": 0.6069,
+      "step": 3602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1482466454767337,
+      "learning_rate": 1.690891392995707e-05,
+      "loss": 0.5591,
+      "step": 3603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2779331072552105,
+      "learning_rate": 1.6907097154805162e-05,
+      "loss": 0.6097,
+      "step": 3604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1215292054998223,
+      "learning_rate": 1.690527994357258e-05,
+      "loss": 0.5335,
+      "step": 3605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2097671267358798,
+      "learning_rate": 1.6903462296374048e-05,
+      "loss": 0.5955,
+      "step": 3606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.265338205880517,
+      "learning_rate": 1.690164421332432e-05,
+      "loss": 0.6472,
+      "step": 3607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1631976505570645,
+      "learning_rate": 1.689982569453819e-05,
+      "loss": 0.6159,
+      "step": 3608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.139486554701297,
+      "learning_rate": 1.689800674013046e-05,
+      "loss": 0.5496,
+      "step": 3609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2219905808362046,
+      "learning_rate": 1.6896187350215977e-05,
+      "loss": 0.5765,
+      "step": 3610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.128260021636023,
+      "learning_rate": 1.689436752490961e-05,
+      "loss": 0.5796,
+      "step": 3611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1476668282703644,
+      "learning_rate": 1.689254726432625e-05,
+      "loss": 0.5473,
+      "step": 3612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2214486492947931,
+      "learning_rate": 1.689072656858082e-05,
+      "loss": 0.5615,
+      "step": 3613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2743865145340358,
+      "learning_rate": 1.6888905437788268e-05,
+      "loss": 0.6311,
+      "step": 3614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1860746060654277,
+      "learning_rate": 1.6887083872063574e-05,
+      "loss": 0.5759,
+      "step": 3615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2199671120979143,
+      "learning_rate": 1.6885261871521746e-05,
+      "loss": 0.6144,
+      "step": 3616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1535508380066237,
+      "learning_rate": 1.688343943627781e-05,
+      "loss": 0.5686,
+      "step": 3617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1471007995347282,
+      "learning_rate": 1.6881616566446827e-05,
+      "loss": 0.5784,
+      "step": 3618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2844511376156666,
+      "learning_rate": 1.6879793262143888e-05,
+      "loss": 0.6096,
+      "step": 3619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2784397661016051,
+      "learning_rate": 1.6877969523484107e-05,
+      "loss": 0.6085,
+      "step": 3620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2485378774076965,
+      "learning_rate": 1.6876145350582623e-05,
+      "loss": 0.6199,
+      "step": 3621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4544783821951675,
+      "learning_rate": 1.6874320743554605e-05,
+      "loss": 0.6546,
+      "step": 3622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.262795192842514,
+      "learning_rate": 1.6872495702515253e-05,
+      "loss": 0.6071,
+      "step": 3623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1761441098587606,
+      "learning_rate": 1.6870670227579788e-05,
+      "loss": 0.6048,
+      "step": 3624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1147398912989954,
+      "learning_rate": 1.6868844318863466e-05,
+      "loss": 0.5831,
+      "step": 3625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2985301199332893,
+      "learning_rate": 1.6867017976481563e-05,
+      "loss": 0.6552,
+      "step": 3626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1591833897108108,
+      "learning_rate": 1.6865191200549387e-05,
+      "loss": 0.619,
+      "step": 3627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3361089049911703,
+      "learning_rate": 1.686336399118227e-05,
+      "loss": 0.6389,
+      "step": 3628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1874962856836957,
+      "learning_rate": 1.686153634849557e-05,
+      "loss": 0.5827,
+      "step": 3629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1986359413308774,
+      "learning_rate": 1.6859708272604685e-05,
+      "loss": 0.5449,
+      "step": 3630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1426136500386732,
+      "learning_rate": 1.6857879763625023e-05,
+      "loss": 0.5697,
+      "step": 3631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2424993543412126,
+      "learning_rate": 1.6856050821672028e-05,
+      "loss": 0.6041,
+      "step": 3632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.279995187511934,
+      "learning_rate": 1.6854221446861175e-05,
+      "loss": 0.6209,
+      "step": 3633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2880656694114747,
+      "learning_rate": 1.6852391639307956e-05,
+      "loss": 0.5896,
+      "step": 3634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1408451860665263,
+      "learning_rate": 1.6850561399127902e-05,
+      "loss": 0.5795,
+      "step": 3635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2170524512454841,
+      "learning_rate": 1.6848730726436562e-05,
+      "loss": 0.6073,
+      "step": 3636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0355150860397122,
+      "learning_rate": 1.6846899621349516e-05,
+      "loss": 0.5646,
+      "step": 3637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.245254042341054,
+      "learning_rate": 1.6845068083982373e-05,
+      "loss": 0.6033,
+      "step": 3638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0928416158426182,
+      "learning_rate": 1.684323611445076e-05,
+      "loss": 0.5769,
+      "step": 3639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1603488954835428,
+      "learning_rate": 1.684140371287035e-05,
+      "loss": 0.5192,
+      "step": 3640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.26295590053536,
+      "learning_rate": 1.6839570879356827e-05,
+      "loss": 0.6128,
+      "step": 3641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0958204019651383,
+      "learning_rate": 1.6837737614025904e-05,
+      "loss": 0.5739,
+      "step": 3642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.138981929069691,
+      "learning_rate": 1.683590391699333e-05,
+      "loss": 0.586,
+      "step": 3643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2919229386528515,
+      "learning_rate": 1.683406978837487e-05,
+      "loss": 0.5916,
+      "step": 3644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2962799085736711,
+      "learning_rate": 1.683223522828633e-05,
+      "loss": 0.6505,
+      "step": 3645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0693247337276943,
+      "learning_rate": 1.6830400236843525e-05,
+      "loss": 0.5842,
+      "step": 3646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2289810148220723,
+      "learning_rate": 1.6828564814162318e-05,
+      "loss": 0.5976,
+      "step": 3647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1618429439528235,
+      "learning_rate": 1.682672896035858e-05,
+      "loss": 0.5479,
+      "step": 3648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1851398705775793,
+      "learning_rate": 1.682489267554822e-05,
+      "loss": 0.6044,
+      "step": 3649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.190137994843335,
+      "learning_rate": 1.6823055959847177e-05,
+      "loss": 0.5815,
+      "step": 3650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.242280968562384,
+      "learning_rate": 1.6821218813371407e-05,
+      "loss": 0.5952,
+      "step": 3651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2239719417167219,
+      "learning_rate": 1.68193812362369e-05,
+      "loss": 0.6235,
+      "step": 3652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.249072780037141,
+      "learning_rate": 1.6817543228559675e-05,
+      "loss": 0.6111,
+      "step": 3653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.255633344228541,
+      "learning_rate": 1.6815704790455768e-05,
+      "loss": 0.5582,
+      "step": 3654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1690479510762544,
+      "learning_rate": 1.6813865922041258e-05,
+      "loss": 0.6173,
+      "step": 3655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1983124231274709,
+      "learning_rate": 1.6812026623432233e-05,
+      "loss": 0.5681,
+      "step": 3656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1530439932470744,
+      "learning_rate": 1.6810186894744825e-05,
+      "loss": 0.5886,
+      "step": 3657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1777359438960062,
+      "learning_rate": 1.680834673609518e-05,
+      "loss": 0.6041,
+      "step": 3658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2249689137154018,
+      "learning_rate": 1.680650614759948e-05,
+      "loss": 0.6975,
+      "step": 3659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2726232815963103,
+      "learning_rate": 1.6804665129373928e-05,
+      "loss": 0.583,
+      "step": 3660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1454927747541033,
+      "learning_rate": 1.6802823681534765e-05,
+      "loss": 0.5201,
+      "step": 3661
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1847175324260097,
+      "learning_rate": 1.680098180419824e-05,
+      "loss": 0.587,
+      "step": 3662
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.212679489850876,
+      "learning_rate": 1.6799139497480644e-05,
+      "loss": 0.6453,
+      "step": 3663
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1576328256579598,
+      "learning_rate": 1.6797296761498295e-05,
+      "loss": 0.5263,
+      "step": 3664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2011842804555024,
+      "learning_rate": 1.6795453596367533e-05,
+      "loss": 0.6175,
+      "step": 3665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2626533471824568,
+      "learning_rate": 1.6793610002204724e-05,
+      "loss": 0.6154,
+      "step": 3666
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2283046037807088,
+      "learning_rate": 1.6791765979126267e-05,
+      "loss": 0.6373,
+      "step": 3667
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1343824885517877,
+      "learning_rate": 1.6789921527248578e-05,
+      "loss": 0.5828,
+      "step": 3668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1749861046801389,
+      "learning_rate": 1.6788076646688117e-05,
+      "loss": 0.6208,
+      "step": 3669
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1817832167161415,
+      "learning_rate": 1.6786231337561352e-05,
+      "loss": 0.5823,
+      "step": 3670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2651603222452013,
+      "learning_rate": 1.6784385599984794e-05,
+      "loss": 0.6719,
+      "step": 3671
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.241392012863191,
+      "learning_rate": 1.678253943407497e-05,
+      "loss": 0.566,
+      "step": 3672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2266400063682432,
+      "learning_rate": 1.6780692839948433e-05,
+      "loss": 0.6127,
+      "step": 3673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2001393912737548,
+      "learning_rate": 1.6778845817721778e-05,
+      "loss": 0.5785,
+      "step": 3674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2902381391063253,
+      "learning_rate": 1.677699836751161e-05,
+      "loss": 0.6447,
+      "step": 3675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.23759219856898,
+      "learning_rate": 1.677515048943457e-05,
+      "loss": 0.5944,
+      "step": 3676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2434291751695066,
+      "learning_rate": 1.6773302183607327e-05,
+      "loss": 0.6044,
+      "step": 3677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1192022443121712,
+      "learning_rate": 1.6771453450146568e-05,
+      "loss": 0.58,
+      "step": 3678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1291153657996924,
+      "learning_rate": 1.676960428916902e-05,
+      "loss": 0.5502,
+      "step": 3679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1483491617079862,
+      "learning_rate": 1.6767754700791425e-05,
+      "loss": 0.5985,
+      "step": 3680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3670887938699179,
+      "learning_rate": 1.676590468513056e-05,
+      "loss": 0.7028,
+      "step": 3681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.197612270436084,
+      "learning_rate": 1.6764054242303223e-05,
+      "loss": 0.5851,
+      "step": 3682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1398866432282182,
+      "learning_rate": 1.6762203372426243e-05,
+      "loss": 0.5857,
+      "step": 3683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.248080496417242,
+      "learning_rate": 1.6760352075616476e-05,
+      "loss": 0.5475,
+      "step": 3684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2040667625528025,
+      "learning_rate": 1.6758500351990808e-05,
+      "loss": 0.6096,
+      "step": 3685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2591083083313643,
+      "learning_rate": 1.675664820166614e-05,
+      "loss": 0.6257,
+      "step": 3686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.265467410225235,
+      "learning_rate": 1.6754795624759414e-05,
+      "loss": 0.5845,
+      "step": 3687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1726670195449058,
+      "learning_rate": 1.6752942621387583e-05,
+      "loss": 0.605,
+      "step": 3688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.303488357345625,
+      "learning_rate": 1.6751089191667648e-05,
+      "loss": 0.5594,
+      "step": 3689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0873268625400927,
+      "learning_rate": 1.6749235335716624e-05,
+      "loss": 0.6074,
+      "step": 3690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2006165748312496,
+      "learning_rate": 1.674738105365155e-05,
+      "loss": 0.5984,
+      "step": 3691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1164834531727266,
+      "learning_rate": 1.67455263455895e-05,
+      "loss": 0.593,
+      "step": 3692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1351784564141443,
+      "learning_rate": 1.6743671211647564e-05,
+      "loss": 0.5658,
+      "step": 3693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1342483891220712,
+      "learning_rate": 1.6741815651942873e-05,
+      "loss": 0.5802,
+      "step": 3694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1397256833189546,
+      "learning_rate": 1.673995966659258e-05,
+      "loss": 0.5906,
+      "step": 3695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3077400725258972,
+      "learning_rate": 1.673810325571386e-05,
+      "loss": 0.6365,
+      "step": 3696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2617434094738549,
+      "learning_rate": 1.6736246419423915e-05,
+      "loss": 0.661,
+      "step": 3697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0985808364441236,
+      "learning_rate": 1.6734389157839975e-05,
+      "loss": 0.5632,
+      "step": 3698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2235864385416373,
+      "learning_rate": 1.673253147107931e-05,
+      "loss": 0.6572,
+      "step": 3699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2304493251659951,
+      "learning_rate": 1.6730673359259194e-05,
+      "loss": 0.618,
+      "step": 3700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1962495010996768,
+      "learning_rate": 1.6728814822496944e-05,
+      "loss": 0.5747,
+      "step": 3701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2330223108022649,
+      "learning_rate": 1.6726955860909903e-05,
+      "loss": 0.5644,
+      "step": 3702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.186116717923621,
+      "learning_rate": 1.6725096474615423e-05,
+      "loss": 0.5529,
+      "step": 3703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1317288707684572,
+      "learning_rate": 1.6723236663730912e-05,
+      "loss": 0.5652,
+      "step": 3704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1843244858892736,
+      "learning_rate": 1.672137642837378e-05,
+      "loss": 0.6114,
+      "step": 3705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2262129558913495,
+      "learning_rate": 1.6719515768661477e-05,
+      "loss": 0.6509,
+      "step": 3706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2056599948010467,
+      "learning_rate": 1.6717654684711475e-05,
+      "loss": 0.6074,
+      "step": 3707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1820732396667126,
+      "learning_rate": 1.6715793176641275e-05,
+      "loss": 0.6061,
+      "step": 3708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.096374034324032,
+      "learning_rate": 1.67139312445684e-05,
+      "loss": 0.5708,
+      "step": 3709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0450719422577202,
+      "learning_rate": 1.671206888861041e-05,
+      "loss": 0.5936,
+      "step": 3710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2613089164530626,
+      "learning_rate": 1.6710206108884884e-05,
+      "loss": 0.5955,
+      "step": 3711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2456495874622497,
+      "learning_rate": 1.6708342905509424e-05,
+      "loss": 0.6818,
+      "step": 3712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2108561826905424,
+      "learning_rate": 1.670647927860166e-05,
+      "loss": 0.5932,
+      "step": 3713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.344744558220863,
+      "learning_rate": 1.6704615228279262e-05,
+      "loss": 0.6318,
+      "step": 3714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.22814940799599,
+      "learning_rate": 1.6702750754659917e-05,
+      "loss": 0.6323,
+      "step": 3715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4351414114362013,
+      "learning_rate": 1.6700885857861332e-05,
+      "loss": 0.6453,
+      "step": 3716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.204848899864932,
+      "learning_rate": 1.6699020538001252e-05,
+      "loss": 0.533,
+      "step": 3717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3624670444667628,
+      "learning_rate": 1.6697154795197442e-05,
+      "loss": 0.6595,
+      "step": 3718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2947579534698466,
+      "learning_rate": 1.6695288629567694e-05,
+      "loss": 0.5823,
+      "step": 3719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1919116839808113,
+      "learning_rate": 1.6693422041229836e-05,
+      "loss": 0.6098,
+      "step": 3720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2553045729257943,
+      "learning_rate": 1.669155503030171e-05,
+      "loss": 0.5954,
+      "step": 3721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.177709576043585,
+      "learning_rate": 1.6689687596901192e-05,
+      "loss": 0.5744,
+      "step": 3722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2484855537571784,
+      "learning_rate": 1.668781974114618e-05,
+      "loss": 0.6338,
+      "step": 3723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1868125029667411,
+      "learning_rate": 1.6685951463154602e-05,
+      "loss": 0.5801,
+      "step": 3724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2722151336629768,
+      "learning_rate": 1.6684082763044415e-05,
+      "loss": 0.6033,
+      "step": 3725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2374335434919408,
+      "learning_rate": 1.66822136409336e-05,
+      "loss": 0.6073,
+      "step": 3726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.354431855082625,
+      "learning_rate": 1.6680344096940157e-05,
+      "loss": 0.6317,
+      "step": 3727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.164652476478852,
+      "learning_rate": 1.667847413118213e-05,
+      "loss": 0.5498,
+      "step": 3728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0840897792398052,
+      "learning_rate": 1.667660374377757e-05,
+      "loss": 0.5425,
+      "step": 3729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2490513539717667,
+      "learning_rate": 1.6674732934844574e-05,
+      "loss": 0.6148,
+      "step": 3730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1622798506131633,
+      "learning_rate": 1.6672861704501247e-05,
+      "loss": 0.5757,
+      "step": 3731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2165724055908238,
+      "learning_rate": 1.6670990052865738e-05,
+      "loss": 0.5865,
+      "step": 3732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2117058070147622,
+      "learning_rate": 1.66691179800562e-05,
+      "loss": 0.5735,
+      "step": 3733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1343691423753561,
+      "learning_rate": 1.6667245486190845e-05,
+      "loss": 0.574,
+      "step": 3734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2180225940325038,
+      "learning_rate": 1.6665372571387882e-05,
+      "loss": 0.5949,
+      "step": 3735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3253238084378547,
+      "learning_rate": 1.6663499235765557e-05,
+      "loss": 0.6714,
+      "step": 3736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2749582583008534,
+      "learning_rate": 1.6661625479442147e-05,
+      "loss": 0.5866,
+      "step": 3737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2200862210714658,
+      "learning_rate": 1.6659751302535952e-05,
+      "loss": 0.6255,
+      "step": 3738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.097326045751031,
+      "learning_rate": 1.6657876705165296e-05,
+      "loss": 0.5646,
+      "step": 3739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1692795558454008,
+      "learning_rate": 1.6656001687448532e-05,
+      "loss": 0.5607,
+      "step": 3740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1575621299465007,
+      "learning_rate": 1.6654126249504042e-05,
+      "loss": 0.6271,
+      "step": 3741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.203734998758077,
+      "learning_rate": 1.665225039145024e-05,
+      "loss": 0.6097,
+      "step": 3742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2704827601596627,
+      "learning_rate": 1.6650374113405536e-05,
+      "loss": 0.6201,
+      "step": 3743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2017453693184021,
+      "learning_rate": 1.664849741548841e-05,
+      "loss": 0.5835,
+      "step": 3744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.176355266178128,
+      "learning_rate": 1.664662029781734e-05,
+      "loss": 0.5748,
+      "step": 3745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.287058573112449,
+      "learning_rate": 1.6644742760510837e-05,
+      "loss": 0.6824,
+      "step": 3746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.13325965211748,
+      "learning_rate": 1.6642864803687443e-05,
+      "loss": 0.5917,
+      "step": 3747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1196763385852007,
+      "learning_rate": 1.664098642746572e-05,
+      "loss": 0.5421,
+      "step": 3748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1415380194268863,
+      "learning_rate": 1.663910763196426e-05,
+      "loss": 0.6158,
+      "step": 3749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.157947994648923,
+      "learning_rate": 1.663722841730168e-05,
+      "loss": 0.5582,
+      "step": 3750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1579000710177414,
+      "learning_rate": 1.663534878359663e-05,
+      "loss": 0.6173,
+      "step": 3751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2256714323703495,
+      "learning_rate": 1.6633468730967778e-05,
+      "loss": 0.625,
+      "step": 3752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2527372906766623,
+      "learning_rate": 1.663158825953382e-05,
+      "loss": 0.6353,
+      "step": 3753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3130623203226444,
+      "learning_rate": 1.662970736941348e-05,
+      "loss": 0.5916,
+      "step": 3754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3642249134113675,
+      "learning_rate": 1.662782606072551e-05,
+      "loss": 0.6803,
+      "step": 3755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3849915399534007,
+      "learning_rate": 1.6625944333588686e-05,
+      "loss": 0.5918,
+      "step": 3756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1508884356706663,
+      "learning_rate": 1.6624062188121808e-05,
+      "loss": 0.5591,
+      "step": 3757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1644823482087894,
+      "learning_rate": 1.662217962444371e-05,
+      "loss": 0.5957,
+      "step": 3758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2755559606713933,
+      "learning_rate": 1.6620296642673248e-05,
+      "loss": 0.6065,
+      "step": 3759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2164102739232026,
+      "learning_rate": 1.6618413242929302e-05,
+      "loss": 0.575,
+      "step": 3760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1921120975992539,
+      "learning_rate": 1.661652942533078e-05,
+      "loss": 0.5584,
+      "step": 3761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0089075807300032,
+      "learning_rate": 1.661464518999662e-05,
+      "loss": 0.5146,
+      "step": 3762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2827660964583378,
+      "learning_rate": 1.6612760537045782e-05,
+      "loss": 0.5962,
+      "step": 3763
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0936471073254517,
+      "learning_rate": 1.6610875466597252e-05,
+      "loss": 0.5608,
+      "step": 3764
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3741402539100864,
+      "learning_rate": 1.660898997877005e-05,
+      "loss": 0.6879,
+      "step": 3765
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0385118904296284,
+      "learning_rate": 1.660710407368321e-05,
+      "loss": 0.5169,
+      "step": 3766
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1351121908355077,
+      "learning_rate": 1.66052177514558e-05,
+      "loss": 0.5849,
+      "step": 3767
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.081335291502842,
+      "learning_rate": 1.660333101220692e-05,
+      "loss": 0.5491,
+      "step": 3768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1485172685181053,
+      "learning_rate": 1.660144385605568e-05,
+      "loss": 0.5717,
+      "step": 3769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2980034296581175,
+      "learning_rate": 1.659955628312123e-05,
+      "loss": 0.6213,
+      "step": 3770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1887873648337839,
+      "learning_rate": 1.6597668293522745e-05,
+      "loss": 0.6154,
+      "step": 3771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1951785043646104,
+      "learning_rate": 1.659577988737942e-05,
+      "loss": 0.6031,
+      "step": 3772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1354199622336807,
+      "learning_rate": 1.659389106481048e-05,
+      "loss": 0.5819,
+      "step": 3773
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2259338128362396,
+      "learning_rate": 1.659200182593518e-05,
+      "loss": 0.5552,
+      "step": 3774
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1859512267995729,
+      "learning_rate": 1.6590112170872792e-05,
+      "loss": 0.5845,
+      "step": 3775
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2402792134026883,
+      "learning_rate": 1.6588222099742624e-05,
+      "loss": 0.5888,
+      "step": 3776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1730735244797954,
+      "learning_rate": 1.6586331612664005e-05,
+      "loss": 0.5874,
+      "step": 3777
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2310936688160352,
+      "learning_rate": 1.658444070975629e-05,
+      "loss": 0.6599,
+      "step": 3778
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2799764677209153,
+      "learning_rate": 1.658254939113886e-05,
+      "loss": 0.6245,
+      "step": 3779
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0665915192021374,
+      "learning_rate": 1.658065765693112e-05,
+      "loss": 0.5781,
+      "step": 3780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1837711516024327,
+      "learning_rate": 1.657876550725252e-05,
+      "loss": 0.6121,
+      "step": 3781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2186858820555029,
+      "learning_rate": 1.6576872942222504e-05,
+      "loss": 0.6413,
+      "step": 3782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2648061293154886,
+      "learning_rate": 1.6574979961960572e-05,
+      "loss": 0.5985,
+      "step": 3783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.221544143878446,
+      "learning_rate": 1.657308656658623e-05,
+      "loss": 0.6057,
+      "step": 3784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.122637758786287,
+      "learning_rate": 1.6571192756219024e-05,
+      "loss": 0.567,
+      "step": 3785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2191033462452205,
+      "learning_rate": 1.6569298530978516e-05,
+      "loss": 0.6446,
+      "step": 3786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2007567642088104,
+      "learning_rate": 1.6567403890984292e-05,
+      "loss": 0.6187,
+      "step": 3787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1891290132263785,
+      "learning_rate": 1.6565508836355983e-05,
+      "loss": 0.5518,
+      "step": 3788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.25788282114907,
+      "learning_rate": 1.6563613367213225e-05,
+      "loss": 0.5823,
+      "step": 3789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2776215044232617,
+      "learning_rate": 1.6561717483675695e-05,
+      "loss": 0.6181,
+      "step": 3790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2788741716522585,
+      "learning_rate": 1.6559821185863082e-05,
+      "loss": 0.5985,
+      "step": 3791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2411446662456593,
+      "learning_rate": 1.6557924473895115e-05,
+      "loss": 0.6061,
+      "step": 3792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2374640815993796,
+      "learning_rate": 1.6556027347891542e-05,
+      "loss": 0.5734,
+      "step": 3793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.346408963394928,
+      "learning_rate": 1.6554129807972135e-05,
+      "loss": 0.6389,
+      "step": 3794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1061710243267628,
+      "learning_rate": 1.6552231854256704e-05,
+      "loss": 0.5716,
+      "step": 3795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2882410379753582,
+      "learning_rate": 1.6550333486865068e-05,
+      "loss": 0.6523,
+      "step": 3796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2567514244422102,
+      "learning_rate": 1.654843470591708e-05,
+      "loss": 0.6178,
+      "step": 3797
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1786103035761748,
+      "learning_rate": 1.654653551153263e-05,
+      "loss": 0.5534,
+      "step": 3798
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1545824965867795,
+      "learning_rate": 1.6544635903831616e-05,
+      "loss": 0.6195,
+      "step": 3799
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1544990685196097,
+      "learning_rate": 1.6542735882933967e-05,
+      "loss": 0.612,
+      "step": 3800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.218692876015038,
+      "learning_rate": 1.6540835448959648e-05,
+      "loss": 0.559,
+      "step": 3801
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2635524404411844,
+      "learning_rate": 1.653893460202864e-05,
+      "loss": 0.6071,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2577346398559268,
+      "learning_rate": 1.6537033342260957e-05,
+      "loss": 0.6566,
+      "step": 3803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1768399664981797,
+      "learning_rate": 1.653513166977663e-05,
+      "loss": 0.596,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1698678693581561,
+      "learning_rate": 1.6533229584695726e-05,
+      "loss": 0.5922,
+      "step": 3805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2614141515821948,
+      "learning_rate": 1.653132708713833e-05,
+      "loss": 0.6109,
+      "step": 3806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3486700253114035,
+      "learning_rate": 1.6529424177224558e-05,
+      "loss": 0.6106,
+      "step": 3807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3491381472580792,
+      "learning_rate": 1.652752085507455e-05,
+      "loss": 0.6157,
+      "step": 3808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1173841930181507,
+      "learning_rate": 1.6525617120808474e-05,
+      "loss": 0.5551,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.181219205505385,
+      "learning_rate": 1.6523712974546522e-05,
+      "loss": 0.6224,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0870230209699292,
+      "learning_rate": 1.652180841640891e-05,
+      "loss": 0.5827,
+      "step": 3811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1779341646848465,
+      "learning_rate": 1.6519903446515884e-05,
+      "loss": 0.6034,
+      "step": 3812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2670920076149634,
+      "learning_rate": 1.6517998064987713e-05,
+      "loss": 0.5926,
+      "step": 3813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1594448284497245,
+      "learning_rate": 1.6516092271944702e-05,
+      "loss": 0.5928,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2274106731659973,
+      "learning_rate": 1.6514186067507164e-05,
+      "loss": 0.6147,
+      "step": 3815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.307674164559868,
+      "learning_rate": 1.651227945179545e-05,
+      "loss": 0.6301,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0382097803781347,
+      "learning_rate": 1.6510372424929938e-05,
+      "loss": 0.5734,
+      "step": 3817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2147871292510144,
+      "learning_rate": 1.650846498703102e-05,
+      "loss": 0.6246,
+      "step": 3818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1487076110822814,
+      "learning_rate": 1.6506557138219136e-05,
+      "loss": 0.5224,
+      "step": 3819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3060120123943155,
+      "learning_rate": 1.6504648878614726e-05,
+      "loss": 0.5835,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0709540678801464,
+      "learning_rate": 1.6502740208338273e-05,
+      "loss": 0.5895,
+      "step": 3821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2362763934890642,
+      "learning_rate": 1.650083112751028e-05,
+      "loss": 0.6449,
+      "step": 3822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1390627783839091,
+      "learning_rate": 1.6498921636251278e-05,
+      "loss": 0.5832,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.10270155865618,
+      "learning_rate": 1.6497011734681824e-05,
+      "loss": 0.5772,
+      "step": 3824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3061348202679655,
+      "learning_rate": 1.64951014229225e-05,
+      "loss": 0.6041,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.150218024525958,
+      "learning_rate": 1.6493190701093913e-05,
+      "loss": 0.5507,
+      "step": 3826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1654001525338058,
+      "learning_rate": 1.6491279569316697e-05,
+      "loss": 0.5608,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2742798248566691,
+      "learning_rate": 1.6489368027711507e-05,
+      "loss": 0.5654,
+      "step": 3828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2060025964522365,
+      "learning_rate": 1.648745607639904e-05,
+      "loss": 0.6007,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2790790867516384,
+      "learning_rate": 1.6485543715499994e-05,
+      "loss": 0.6137,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.211047505950281,
+      "learning_rate": 1.6483630945135112e-05,
+      "loss": 0.632,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1557007077692802,
+      "learning_rate": 1.648171776542516e-05,
+      "loss": 0.5673,
+      "step": 3832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1948993442552436,
+      "learning_rate": 1.647980417649092e-05,
+      "loss": 0.6027,
+      "step": 3833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.250139705480305,
+      "learning_rate": 1.6477890178453216e-05,
+      "loss": 0.5999,
+      "step": 3834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0587319146675518,
+      "learning_rate": 1.6475975771432883e-05,
+      "loss": 0.5594,
+      "step": 3835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1433898129336177,
+      "learning_rate": 1.6474060955550783e-05,
+      "loss": 0.5648,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1182726411699961,
+      "learning_rate": 1.647214573092782e-05,
+      "loss": 0.5938,
+      "step": 3837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2268435891124432,
+      "learning_rate": 1.64702300976849e-05,
+      "loss": 0.656,
+      "step": 3838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1606531074666053,
+      "learning_rate": 1.6468314055942977e-05,
+      "loss": 0.5799,
+      "step": 3839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1225013535795143,
+      "learning_rate": 1.646639760582301e-05,
+      "loss": 0.6052,
+      "step": 3840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3084316324048235,
+      "learning_rate": 1.6464480747446e-05,
+      "loss": 0.6426,
+      "step": 3841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2054317209090708,
+      "learning_rate": 1.646256348093297e-05,
+      "loss": 0.5734,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2244733768371143,
+      "learning_rate": 1.6460645806404967e-05,
+      "loss": 0.5561,
+      "step": 3843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.147653221090019,
+      "learning_rate": 1.645872772398306e-05,
+      "loss": 0.6205,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1980938932373515,
+      "learning_rate": 1.645680923378835e-05,
+      "loss": 0.5896,
+      "step": 3845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0626539231050376,
+      "learning_rate": 1.6454890335941957e-05,
+      "loss": 0.5643,
+      "step": 3846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2680301653736579,
+      "learning_rate": 1.645297103056504e-05,
+      "loss": 0.6094,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.213017946892014,
+      "learning_rate": 1.6451051317778764e-05,
+      "loss": 0.6437,
+      "step": 3848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.265063502711842,
+      "learning_rate": 1.6449131197704337e-05,
+      "loss": 0.6289,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.218225366206749,
+      "learning_rate": 1.6447210670462985e-05,
+      "loss": 0.636,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.20334291342482,
+      "learning_rate": 1.644528973617596e-05,
+      "loss": 0.6317,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1452117042892074,
+      "learning_rate": 1.644336839496454e-05,
+      "loss": 0.5835,
+      "step": 3852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3163676270385396,
+      "learning_rate": 1.6441446646950027e-05,
+      "loss": 0.6678,
+      "step": 3853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1686447891777894,
+      "learning_rate": 1.643952449225376e-05,
+      "loss": 0.5849,
+      "step": 3854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1595691261330412,
+      "learning_rate": 1.6437601930997083e-05,
+      "loss": 0.5801,
+      "step": 3855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2232351181993006,
+      "learning_rate": 1.643567896330138e-05,
+      "loss": 0.5258,
+      "step": 3856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.235197300382614,
+      "learning_rate": 1.6433755589288067e-05,
+      "loss": 0.6101,
+      "step": 3857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1154672142453195,
+      "learning_rate": 1.643183180907857e-05,
+      "loss": 0.5405,
+      "step": 3858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.10278393279012,
+      "learning_rate": 1.642990762279434e-05,
+      "loss": 0.5437,
+      "step": 3859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.223961958643096,
+      "learning_rate": 1.6427983030556872e-05,
+      "loss": 0.6144,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3024641268543844,
+      "learning_rate": 1.6426058032487675e-05,
+      "loss": 0.6771,
+      "step": 3861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1650502667204066,
+      "learning_rate": 1.6424132628708276e-05,
+      "loss": 0.6146,
+      "step": 3862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1345842909251438,
+      "learning_rate": 1.642220681934024e-05,
+      "loss": 0.55,
+      "step": 3863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1848294697157156,
+      "learning_rate": 1.6420280604505152e-05,
+      "loss": 0.543,
+      "step": 3864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2912866074283447,
+      "learning_rate": 1.6418353984324628e-05,
+      "loss": 0.657,
+      "step": 3865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1422981114715618,
+      "learning_rate": 1.6416426958920303e-05,
+      "loss": 0.612,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1744473009657208,
+      "learning_rate": 1.6414499528413838e-05,
+      "loss": 0.5618,
+      "step": 3867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1696305206713205,
+      "learning_rate": 1.6412571692926924e-05,
+      "loss": 0.5347,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2394894262480667,
+      "learning_rate": 1.6410643452581274e-05,
+      "loss": 0.5943,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1992516528657549,
+      "learning_rate": 1.640871480749863e-05,
+      "loss": 0.6021,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2168749543882105,
+      "learning_rate": 1.6406785757800753e-05,
+      "loss": 0.6033,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2110106907537241,
+      "learning_rate": 1.6404856303609435e-05,
+      "loss": 0.6042,
+      "step": 3872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0960628034773212,
+      "learning_rate": 1.64029264450465e-05,
+      "loss": 0.5802,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1151483238679345,
+      "learning_rate": 1.640099618223378e-05,
+      "loss": 0.601,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2044190222668023,
+      "learning_rate": 1.6399065515293145e-05,
+      "loss": 0.6263,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1307449414690194,
+      "learning_rate": 1.6397134444346487e-05,
+      "loss": 0.5414,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.253980640331903,
+      "learning_rate": 1.6395202969515735e-05,
+      "loss": 0.6337,
+      "step": 3877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2784498832436724,
+      "learning_rate": 1.639327109092282e-05,
+      "loss": 0.614,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3343225723437493,
+      "learning_rate": 1.639133880868971e-05,
+      "loss": 0.6371,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1959331123026353,
+      "learning_rate": 1.6389406122938415e-05,
+      "loss": 0.5897,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.174967385914844,
+      "learning_rate": 1.6387473033790944e-05,
+      "loss": 0.5861,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1046235320955742,
+      "learning_rate": 1.6385539541369346e-05,
+      "loss": 0.551,
+      "step": 3882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2528323033842708,
+      "learning_rate": 1.6383605645795687e-05,
+      "loss": 0.6035,
+      "step": 3883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2644852580836072,
+      "learning_rate": 1.6381671347192073e-05,
+      "loss": 0.5991,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2052585955101447,
+      "learning_rate": 1.6379736645680622e-05,
+      "loss": 0.561,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3026870650318307,
+      "learning_rate": 1.6377801541383477e-05,
+      "loss": 0.6556,
+      "step": 3886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.350319042300315,
+      "learning_rate": 1.637586603442282e-05,
+      "loss": 0.6403,
+      "step": 3887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1535409688829779,
+      "learning_rate": 1.6373930124920844e-05,
+      "loss": 0.5629,
+      "step": 3888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.300501337950479,
+      "learning_rate": 1.6371993812999773e-05,
+      "loss": 0.5918,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4052594829022873,
+      "learning_rate": 1.637005709878186e-05,
+      "loss": 0.6102,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2750333370263558,
+      "learning_rate": 1.6368119982389373e-05,
+      "loss": 0.6711,
+      "step": 3891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2093439260822887,
+      "learning_rate": 1.6366182463944618e-05,
+      "loss": 0.6573,
+      "step": 3892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2163617625152745,
+      "learning_rate": 1.636424454356992e-05,
+      "loss": 0.5821,
+      "step": 3893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1146755120969853,
+      "learning_rate": 1.636230622138763e-05,
+      "loss": 0.6064,
+      "step": 3894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.179718977148219,
+      "learning_rate": 1.636036749752012e-05,
+      "loss": 0.5514,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3208943726949323,
+      "learning_rate": 1.63584283720898e-05,
+      "loss": 0.6838,
+      "step": 3896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3103264567866675,
+      "learning_rate": 1.6356488845219086e-05,
+      "loss": 0.5689,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3054659228748,
+      "learning_rate": 1.6354548917030437e-05,
+      "loss": 0.5876,
+      "step": 3898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1243838106226864,
+      "learning_rate": 1.635260858764633e-05,
+      "loss": 0.5186,
+      "step": 3899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2194864542209984,
+      "learning_rate": 1.6350667857189268e-05,
+      "loss": 0.6127,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.275604744151813,
+      "learning_rate": 1.634872672578178e-05,
+      "loss": 0.6614,
+      "step": 3901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.179660064176127,
+      "learning_rate": 1.6346785193546418e-05,
+      "loss": 0.5828,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1594623584172743,
+      "learning_rate": 1.634484326060576e-05,
+      "loss": 0.5918,
+      "step": 3903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2652293399193577,
+      "learning_rate": 1.6342900927082415e-05,
+      "loss": 0.5914,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2085145891103632,
+      "learning_rate": 1.634095819309901e-05,
+      "loss": 0.5946,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2495752566641438,
+      "learning_rate": 1.6339015058778196e-05,
+      "loss": 0.5987,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3488035321859129,
+      "learning_rate": 1.6337071524242657e-05,
+      "loss": 0.6658,
+      "step": 3907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0699372016446838,
+      "learning_rate": 1.6335127589615097e-05,
+      "loss": 0.5674,
+      "step": 3908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1791559499696191,
+      "learning_rate": 1.6333183255018247e-05,
+      "loss": 0.5647,
+      "step": 3909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2328335275536846,
+      "learning_rate": 1.6331238520574866e-05,
+      "loss": 0.6384,
+      "step": 3910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1918939311499923,
+      "learning_rate": 1.632929338640773e-05,
+      "loss": 0.6079,
+      "step": 3911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3075585213250527,
+      "learning_rate": 1.632734785263965e-05,
+      "loss": 0.6057,
+      "step": 3912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.167633161527098,
+      "learning_rate": 1.6325401919393455e-05,
+      "loss": 0.5506,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.248550910238651,
+      "learning_rate": 1.6323455586792e-05,
+      "loss": 0.6152,
+      "step": 3914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0655156143741384,
+      "learning_rate": 1.632150885495817e-05,
+      "loss": 0.5355,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1881358803256417,
+      "learning_rate": 1.631956172401487e-05,
+      "loss": 0.6435,
+      "step": 3916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2162741430695418,
+      "learning_rate": 1.6317614194085033e-05,
+      "loss": 0.5917,
+      "step": 3917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5600653945225398,
+      "learning_rate": 1.631566626529162e-05,
+      "loss": 0.5381,
+      "step": 3918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.076173093487003,
+      "learning_rate": 1.6313717937757612e-05,
+      "loss": 0.5388,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1156762845943378,
+      "learning_rate": 1.6311769211606012e-05,
+      "loss": 0.5769,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.27629685737159,
+      "learning_rate": 1.630982008695986e-05,
+      "loss": 0.6195,
+      "step": 3921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2025496853657713,
+      "learning_rate": 1.6307870563942212e-05,
+      "loss": 0.5878,
+      "step": 3922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2169623346635952,
+      "learning_rate": 1.630592064267615e-05,
+      "loss": 0.5994,
+      "step": 3923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2369774533979363,
+      "learning_rate": 1.6303970323284784e-05,
+      "loss": 0.633,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.119677669429823,
+      "learning_rate": 1.630201960589125e-05,
+      "loss": 0.5875,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1193954949816354,
+      "learning_rate": 1.63000684906187e-05,
+      "loss": 0.5378,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1732046594741175,
+      "learning_rate": 1.6298116977590323e-05,
+      "loss": 0.5885,
+      "step": 3927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1824317994723597,
+      "learning_rate": 1.6296165066929327e-05,
+      "loss": 0.5893,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1695816487860884,
+      "learning_rate": 1.6294212758758945e-05,
+      "loss": 0.6354,
+      "step": 3929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1942416611624476,
+      "learning_rate": 1.629226005320244e-05,
+      "loss": 0.6118,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1750110625699934,
+      "learning_rate": 1.6290306950383095e-05,
+      "loss": 0.6372,
+      "step": 3931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2056301837056256,
+      "learning_rate": 1.6288353450424216e-05,
+      "loss": 0.5981,
+      "step": 3932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2117346324162617,
+      "learning_rate": 1.628639955344914e-05,
+      "loss": 0.5877,
+      "step": 3933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.21289696438372,
+      "learning_rate": 1.628444525958123e-05,
+      "loss": 0.6124,
+      "step": 3934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1563761358794684,
+      "learning_rate": 1.6282490568943867e-05,
+      "loss": 0.5792,
+      "step": 3935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.254263996179238,
+      "learning_rate": 1.6280535481660454e-05,
+      "loss": 0.642,
+      "step": 3936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2512793669555577,
+      "learning_rate": 1.627857999785444e-05,
+      "loss": 0.5883,
+      "step": 3937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1928629990067652,
+      "learning_rate": 1.6276624117649273e-05,
+      "loss": 0.6174,
+      "step": 3938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2099964546120145,
+      "learning_rate": 1.6274667841168445e-05,
+      "loss": 0.572,
+      "step": 3939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1618369929277412,
+      "learning_rate": 1.6272711168535465e-05,
+      "loss": 0.6543,
+      "step": 3940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2139831634874838,
+      "learning_rate": 1.6270754099873866e-05,
+      "loss": 0.572,
+      "step": 3941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2067460892991102,
+      "learning_rate": 1.62687966353072e-05,
+      "loss": 0.6673,
+      "step": 3942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1363861541250981,
+      "learning_rate": 1.6266838774959065e-05,
+      "loss": 0.5476,
+      "step": 3943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.159579149559187,
+      "learning_rate": 1.6264880518953068e-05,
+      "loss": 0.5904,
+      "step": 3944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0840394702003568,
+      "learning_rate": 1.626292186741284e-05,
+      "loss": 0.5894,
+      "step": 3945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3137632829918364,
+      "learning_rate": 1.626096282046204e-05,
+      "loss": 0.6124,
+      "step": 3946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.199484929170182,
+      "learning_rate": 1.6259003378224358e-05,
+      "loss": 0.5361,
+      "step": 3947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.126221523454137,
+      "learning_rate": 1.6257043540823498e-05,
+      "loss": 0.5546,
+      "step": 3948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1876042972740064,
+      "learning_rate": 1.6255083308383195e-05,
+      "loss": 0.5831,
+      "step": 3949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0887787239849271,
+      "learning_rate": 1.6253122681027214e-05,
+      "loss": 0.6001,
+      "step": 3950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2446340304149863,
+      "learning_rate": 1.6251161658879338e-05,
+      "loss": 0.6371,
+      "step": 3951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2377869976433886,
+      "learning_rate": 1.6249200242063368e-05,
+      "loss": 0.5563,
+      "step": 3952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1588291065492486,
+      "learning_rate": 1.624723843070315e-05,
+      "loss": 0.638,
+      "step": 3953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.111704462620387,
+      "learning_rate": 1.6245276224922538e-05,
+      "loss": 0.602,
+      "step": 3954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.234620564818056,
+      "learning_rate": 1.6243313624845417e-05,
+      "loss": 0.5974,
+      "step": 3955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2223304582604837,
+      "learning_rate": 1.6241350630595697e-05,
+      "loss": 0.5917,
+      "step": 3956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1517162101364646,
+      "learning_rate": 1.623938724229731e-05,
+      "loss": 0.5855,
+      "step": 3957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.271761909349612,
+      "learning_rate": 1.623742346007421e-05,
+      "loss": 0.6164,
+      "step": 3958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2746847866335282,
+      "learning_rate": 1.6235459284050393e-05,
+      "loss": 0.6274,
+      "step": 3959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2029244887482289,
+      "learning_rate": 1.6233494714349854e-05,
+      "loss": 0.6257,
+      "step": 3960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1819769262524185,
+      "learning_rate": 1.623152975109664e-05,
+      "loss": 0.6301,
+      "step": 3961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2891098476153031,
+      "learning_rate": 1.6229564394414796e-05,
+      "loss": 0.6508,
+      "step": 3962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2121481965142733,
+      "learning_rate": 1.6227598644428417e-05,
+      "loss": 0.5883,
+      "step": 3963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2397736421989687,
+      "learning_rate": 1.6225632501261603e-05,
+      "loss": 0.6003,
+      "step": 3964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.241069074277615,
+      "learning_rate": 1.622366596503849e-05,
+      "loss": 0.5807,
+      "step": 3965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2458192050519732,
+      "learning_rate": 1.622169903588323e-05,
+      "loss": 0.581,
+      "step": 3966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.15965748350523,
+      "learning_rate": 1.6219731713920017e-05,
+      "loss": 0.583,
+      "step": 3967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2442792638549236,
+      "learning_rate": 1.621776399927305e-05,
+      "loss": 0.5626,
+      "step": 3968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1158338762604407,
+      "learning_rate": 1.6215795892066556e-05,
+      "loss": 0.5716,
+      "step": 3969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.216933486152108,
+      "learning_rate": 1.6213827392424802e-05,
+      "loss": 0.5675,
+      "step": 3970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3129663547147643,
+      "learning_rate": 1.6211858500472068e-05,
+      "loss": 0.6168,
+      "step": 3971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5181623632828118,
+      "learning_rate": 1.6209889216332657e-05,
+      "loss": 0.5845,
+      "step": 3972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2594047561802864,
+      "learning_rate": 1.62079195401309e-05,
+      "loss": 0.6096,
+      "step": 3973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1115418883893513,
+      "learning_rate": 1.6205949471991154e-05,
+      "loss": 0.6022,
+      "step": 3974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2668267655000156,
+      "learning_rate": 1.6203979012037797e-05,
+      "loss": 0.6232,
+      "step": 3975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1391563366410193,
+      "learning_rate": 1.620200816039524e-05,
+      "loss": 0.5601,
+      "step": 3976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2115418925487111,
+      "learning_rate": 1.6200036917187908e-05,
+      "loss": 0.6136,
+      "step": 3977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2694049948487522,
+      "learning_rate": 1.6198065282540258e-05,
+      "loss": 0.6167,
+      "step": 3978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2722178510223396,
+      "learning_rate": 1.619609325657677e-05,
+      "loss": 0.6318,
+      "step": 3979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3864098863092906,
+      "learning_rate": 1.619412083942195e-05,
+      "loss": 0.6064,
+      "step": 3980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1347166172284997,
+      "learning_rate": 1.6192148031200315e-05,
+      "loss": 0.5682,
+      "step": 3981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1315261377340633,
+      "learning_rate": 1.6190174832036434e-05,
+      "loss": 0.5535,
+      "step": 3982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.119340436051421,
+      "learning_rate": 1.6188201242054875e-05,
+      "loss": 0.5955,
+      "step": 3983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2052149764493065,
+      "learning_rate": 1.6186227261380247e-05,
+      "loss": 0.6179,
+      "step": 3984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.170635228802279,
+      "learning_rate": 1.6184252890137175e-05,
+      "loss": 0.6167,
+      "step": 3985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1688227768541013,
+      "learning_rate": 1.618227812845031e-05,
+      "loss": 0.5885,
+      "step": 3986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2806564444855968,
+      "learning_rate": 1.6180302976444332e-05,
+      "loss": 0.6695,
+      "step": 3987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1579952471574613,
+      "learning_rate": 1.617832743424394e-05,
+      "loss": 0.5816,
+      "step": 3988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1821707553065879,
+      "learning_rate": 1.6176351501973865e-05,
+      "loss": 0.6079,
+      "step": 3989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.191993693626286,
+      "learning_rate": 1.6174375179758854e-05,
+      "loss": 0.6214,
+      "step": 3990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2450894221781705,
+      "learning_rate": 1.6172398467723674e-05,
+      "loss": 0.5836,
+      "step": 3991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2226072551429032,
+      "learning_rate": 1.617042136599314e-05,
+      "loss": 0.6354,
+      "step": 3992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1979623979325693,
+      "learning_rate": 1.6168443874692066e-05,
+      "loss": 0.5994,
+      "step": 3993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4221004265466641,
+      "learning_rate": 1.616646599394531e-05,
+      "loss": 0.5728,
+      "step": 3994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2873378892337566,
+      "learning_rate": 1.616448772387774e-05,
+      "loss": 0.6059,
+      "step": 3995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.061620235855083,
+      "learning_rate": 1.6162509064614253e-05,
+      "loss": 0.5424,
+      "step": 3996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1789231508331548,
+      "learning_rate": 1.6160530016279774e-05,
+      "loss": 0.4952,
+      "step": 3997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1910281222279882,
+      "learning_rate": 1.6158550578999252e-05,
+      "loss": 0.6121,
+      "step": 3998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.125059867961342,
+      "learning_rate": 1.6156570752897658e-05,
+      "loss": 0.5408,
+      "step": 3999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2594834595730304,
+      "learning_rate": 1.6154590538099993e-05,
+      "loss": 0.5299,
+      "step": 4000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.153355816739578,
+      "learning_rate": 1.6152609934731266e-05,
+      "loss": 0.5875,
+      "step": 4001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.139540431012989,
+      "learning_rate": 1.615062894291654e-05,
+      "loss": 0.5971,
+      "step": 4002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1937285755891243,
+      "learning_rate": 1.6148647562780865e-05,
+      "loss": 0.5594,
+      "step": 4003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1929602821913936,
+      "learning_rate": 1.6146665794449352e-05,
+      "loss": 0.626,
+      "step": 4004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2669126765290302,
+      "learning_rate": 1.6144683638047115e-05,
+      "loss": 0.6363,
+      "step": 4005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1527505964626341,
+      "learning_rate": 1.6142701093699296e-05,
+      "loss": 0.5758,
+      "step": 4006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2021356949181223,
+      "learning_rate": 1.6140718161531066e-05,
+      "loss": 0.5874,
+      "step": 4007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2117869689351568,
+      "learning_rate": 1.613873484166762e-05,
+      "loss": 0.6391,
+      "step": 4008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1044450744780192,
+      "learning_rate": 1.6136751134234163e-05,
+      "loss": 0.5521,
+      "step": 4009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2154085863952253,
+      "learning_rate": 1.6134767039355953e-05,
+      "loss": 0.6514,
+      "step": 4010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.178474914636623,
+      "learning_rate": 1.6132782557158245e-05,
+      "loss": 0.5812,
+      "step": 4011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1189564391888203,
+      "learning_rate": 1.6130797687766335e-05,
+      "loss": 0.544,
+      "step": 4012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1246021944951656,
+      "learning_rate": 1.6128812431305534e-05,
+      "loss": 0.5812,
+      "step": 4013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.175541647390319,
+      "learning_rate": 1.6126826787901185e-05,
+      "loss": 0.6081,
+      "step": 4014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1366678337937643,
+      "learning_rate": 1.6124840757678653e-05,
+      "loss": 0.6106,
+      "step": 4015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1244202815591704,
+      "learning_rate": 1.612285434076332e-05,
+      "loss": 0.5615,
+      "step": 4016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2184845072779211,
+      "learning_rate": 1.6120867537280608e-05,
+      "loss": 0.6,
+      "step": 4017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1816462240795806,
+      "learning_rate": 1.611888034735594e-05,
+      "loss": 0.6147,
+      "step": 4018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2339105818886065,
+      "learning_rate": 1.6116892771114793e-05,
+      "loss": 0.633,
+      "step": 4019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0915302094260146,
+      "learning_rate": 1.6114904808682644e-05,
+      "loss": 0.5481,
+      "step": 4020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1885559506885552,
+      "learning_rate": 1.6112916460185008e-05,
+      "loss": 0.6277,
+      "step": 4021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2741424855810366,
+      "learning_rate": 1.6110927725747413e-05,
+      "loss": 0.6197,
+      "step": 4022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1242694602064187,
+      "learning_rate": 1.610893860549543e-05,
+      "loss": 0.5296,
+      "step": 4023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1756699214710347,
+      "learning_rate": 1.610694909955463e-05,
+      "loss": 0.5201,
+      "step": 4024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2186917022064503,
+      "learning_rate": 1.6104959208050625e-05,
+      "loss": 0.6423,
+      "step": 4025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1961654412370195,
+      "learning_rate": 1.610296893110905e-05,
+      "loss": 0.586,
+      "step": 4026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1615095369601849,
+      "learning_rate": 1.6100978268855552e-05,
+      "loss": 0.5639,
+      "step": 4027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2014112280579905,
+      "learning_rate": 1.609898722141583e-05,
+      "loss": 0.6249,
+      "step": 4028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1315937194478678,
+      "learning_rate": 1.609699578891557e-05,
+      "loss": 0.5543,
+      "step": 4029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3114084064089124,
+      "learning_rate": 1.6095003971480514e-05,
+      "loss": 0.6376,
+      "step": 4030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.265077448950562,
+      "learning_rate": 1.609301176923641e-05,
+      "loss": 0.6216,
+      "step": 4031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1475924022928314,
+      "learning_rate": 1.6091019182309033e-05,
+      "loss": 0.5656,
+      "step": 4032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1737722170903604,
+      "learning_rate": 1.6089026210824196e-05,
+      "loss": 0.6254,
+      "step": 4033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1438404381021985,
+      "learning_rate": 1.6087032854907715e-05,
+      "loss": 0.5805,
+      "step": 4034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0830969674895723,
+      "learning_rate": 1.6085039114685448e-05,
+      "loss": 0.5331,
+      "step": 4035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1128636698004142,
+      "learning_rate": 1.6083044990283263e-05,
+      "loss": 0.5696,
+      "step": 4036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.194368126498595,
+      "learning_rate": 1.6081050481827066e-05,
+      "loss": 0.6,
+      "step": 4037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1239262861197168,
+      "learning_rate": 1.6079055589442778e-05,
+      "loss": 0.6219,
+      "step": 4038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2518657588379594,
+      "learning_rate": 1.6077060313256348e-05,
+      "loss": 0.6307,
+      "step": 4039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.090109953277923,
+      "learning_rate": 1.6075064653393748e-05,
+      "loss": 0.5383,
+      "step": 4040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2473465412676827,
+      "learning_rate": 1.6073068609980968e-05,
+      "loss": 0.5926,
+      "step": 4041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3225680615338957,
+      "learning_rate": 1.6071072183144033e-05,
+      "loss": 0.5962,
+      "step": 4042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2329081257718428,
+      "learning_rate": 1.6069075373008997e-05,
+      "loss": 0.597,
+      "step": 4043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1676300986786838,
+      "learning_rate": 1.6067078179701913e-05,
+      "loss": 0.5808,
+      "step": 4044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1526814628616309,
+      "learning_rate": 1.6065080603348885e-05,
+      "loss": 0.6044,
+      "step": 4045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2083597125539236,
+      "learning_rate": 1.6063082644076026e-05,
+      "loss": 0.6071,
+      "step": 4046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2155312313208602,
+      "learning_rate": 1.6061084302009473e-05,
+      "loss": 0.6456,
+      "step": 4047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1809232189145735,
+      "learning_rate": 1.6059085577275402e-05,
+      "loss": 0.6001,
+      "step": 4048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3228249655752855,
+      "learning_rate": 1.605708647e-05,
+      "loss": 0.5856,
+      "step": 4049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.174252146273254,
+      "learning_rate": 1.6055086980309475e-05,
+      "loss": 0.6324,
+      "step": 4050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.333198326944949,
+      "learning_rate": 1.605308710833007e-05,
+      "loss": 0.5949,
+      "step": 4051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1584392647823685,
+      "learning_rate": 1.6051086854188046e-05,
+      "loss": 0.5239,
+      "step": 4052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.042404193455187,
+      "learning_rate": 1.604908621800969e-05,
+      "loss": 0.5189,
+      "step": 4053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1160018604939856,
+      "learning_rate": 1.604708519992131e-05,
+      "loss": 0.5914,
+      "step": 4054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.246749179390805,
+      "learning_rate": 1.6045083800049243e-05,
+      "loss": 0.5528,
+      "step": 4055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3248094349702548,
+      "learning_rate": 1.604308201851985e-05,
+      "loss": 0.638,
+      "step": 4056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1998209978746488,
+      "learning_rate": 1.6041079855459506e-05,
+      "loss": 0.6108,
+      "step": 4057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2605346225158425,
+      "learning_rate": 1.603907731099463e-05,
+      "loss": 0.5983,
+      "step": 4058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2750747079092335,
+      "learning_rate": 1.603707438525164e-05,
+      "loss": 0.6677,
+      "step": 4059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2016606522690463,
+      "learning_rate": 1.6035071078356998e-05,
+      "loss": 0.5742,
+      "step": 4060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1839374509025111,
+      "learning_rate": 1.603306739043718e-05,
+      "loss": 0.5602,
+      "step": 4061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1035169213211413,
+      "learning_rate": 1.6031063321618697e-05,
+      "loss": 0.5843,
+      "step": 4062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.200744999652961,
+      "learning_rate": 1.6029058872028063e-05,
+      "loss": 0.6201,
+      "step": 4063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.152203101933758,
+      "learning_rate": 1.6027054041791842e-05,
+      "loss": 0.6164,
+      "step": 4064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.256626826083971,
+      "learning_rate": 1.6025048831036603e-05,
+      "loss": 0.638,
+      "step": 4065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1813671457644892,
+      "learning_rate": 1.6023043239888946e-05,
+      "loss": 0.6082,
+      "step": 4066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2103966274123723,
+      "learning_rate": 1.602103726847549e-05,
+      "loss": 0.5989,
+      "step": 4067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2175937939181174,
+      "learning_rate": 1.6019030916922892e-05,
+      "loss": 0.6199,
+      "step": 4068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1165489025553181,
+      "learning_rate": 1.6017024185357817e-05,
+      "loss": 0.5692,
+      "step": 4069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.250109190940164,
+      "learning_rate": 1.601501707390696e-05,
+      "loss": 0.5884,
+      "step": 4070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3070057813462548,
+      "learning_rate": 1.6013009582697043e-05,
+      "loss": 0.6139,
+      "step": 4071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2741008037379273,
+      "learning_rate": 1.601100171185481e-05,
+      "loss": 0.6513,
+      "step": 4072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0050610383320062,
+      "learning_rate": 1.6008993461507022e-05,
+      "loss": 0.4993,
+      "step": 4073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1282934089558068,
+      "learning_rate": 1.6006984831780476e-05,
+      "loss": 0.5376,
+      "step": 4074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2381305783419083,
+      "learning_rate": 1.6004975822801986e-05,
+      "loss": 0.6253,
+      "step": 4075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2529711697877712,
+      "learning_rate": 1.600296643469839e-05,
+      "loss": 0.6211,
+      "step": 4076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.146491861400174,
+      "learning_rate": 1.6000956667596554e-05,
+      "loss": 0.5507,
+      "step": 4077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0868085482674297,
+      "learning_rate": 1.599894652162336e-05,
+      "loss": 0.557,
+      "step": 4078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2270822577911655,
+      "learning_rate": 1.5996935996905722e-05,
+      "loss": 0.5649,
+      "step": 4079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1749864597757853,
+      "learning_rate": 1.5994925093570578e-05,
+      "loss": 0.6118,
+      "step": 4080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2632427164876006,
+      "learning_rate": 1.599291381174488e-05,
+      "loss": 0.5955,
+      "step": 4081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.183447246486196,
+      "learning_rate": 1.5990902151555612e-05,
+      "loss": 0.5786,
+      "step": 4082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.124024498058253,
+      "learning_rate": 1.5988890113129786e-05,
+      "loss": 0.5757,
+      "step": 4083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1925538586303988,
+      "learning_rate": 1.5986877696594425e-05,
+      "loss": 0.6069,
+      "step": 4084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2013483181511755,
+      "learning_rate": 1.598486490207659e-05,
+      "loss": 0.6026,
+      "step": 4085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0765093431789332,
+      "learning_rate": 1.598285172970335e-05,
+      "loss": 0.5528,
+      "step": 4086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2788229027901965,
+      "learning_rate": 1.598083817960182e-05,
+      "loss": 0.6258,
+      "step": 4087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3049319831983297,
+      "learning_rate": 1.5978824251899117e-05,
+      "loss": 0.6535,
+      "step": 4088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2464490043286351,
+      "learning_rate": 1.597680994672239e-05,
+      "loss": 0.6026,
+      "step": 4089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1429909506445404,
+      "learning_rate": 1.5974795264198814e-05,
+      "loss": 0.6242,
+      "step": 4090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2879980605359609,
+      "learning_rate": 1.5972780204455592e-05,
+      "loss": 0.6501,
+      "step": 4091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1755463121475838,
+      "learning_rate": 1.5970764767619933e-05,
+      "loss": 0.5684,
+      "step": 4092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.196094880206311,
+      "learning_rate": 1.5968748953819095e-05,
+      "loss": 0.6099,
+      "step": 4093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2817598584493335,
+      "learning_rate": 1.596673276318034e-05,
+      "loss": 0.5943,
+      "step": 4094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2085586809077369,
+      "learning_rate": 1.5964716195830958e-05,
+      "loss": 0.6354,
+      "step": 4095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.35159720668442,
+      "learning_rate": 1.5962699251898274e-05,
+      "loss": 0.6055,
+      "step": 4096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.133824705209295,
+      "learning_rate": 1.596068193150962e-05,
+      "loss": 0.5596,
+      "step": 4097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1371854368119316,
+      "learning_rate": 1.595866423479236e-05,
+      "loss": 0.5844,
+      "step": 4098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1801612357528344,
+      "learning_rate": 1.595664616187389e-05,
+      "loss": 0.5828,
+      "step": 4099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2468891056955147,
+      "learning_rate": 1.595462771288161e-05,
+      "loss": 0.6108,
+      "step": 4100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2033901170205261,
+      "learning_rate": 1.5952608887942967e-05,
+      "loss": 0.6022,
+      "step": 4101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2634317679962785,
+      "learning_rate": 1.5950589687185405e-05,
+      "loss": 0.6315,
+      "step": 4102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.168779021938064,
+      "learning_rate": 1.594857011073642e-05,
+      "loss": 0.56,
+      "step": 4103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0974620497969982,
+      "learning_rate": 1.5946550158723516e-05,
+      "loss": 0.5621,
+      "step": 4104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0945140349881082,
+      "learning_rate": 1.5944529831274213e-05,
+      "loss": 0.5609,
+      "step": 4105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1688990635594247,
+      "learning_rate": 1.5942509128516077e-05,
+      "loss": 0.5748,
+      "step": 4106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2062149003214337,
+      "learning_rate": 1.594048805057668e-05,
+      "loss": 0.6537,
+      "step": 4107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.290696605330871,
+      "learning_rate": 1.5938466597583625e-05,
+      "loss": 0.6344,
+      "step": 4108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2284935492880682,
+      "learning_rate": 1.5936444769664533e-05,
+      "loss": 0.5891,
+      "step": 4109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2105269197466848,
+      "learning_rate": 1.593442256694705e-05,
+      "loss": 0.631,
+      "step": 4110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.196250198667185,
+      "learning_rate": 1.593239998955886e-05,
+      "loss": 0.593,
+      "step": 4111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.285429487777566,
+      "learning_rate": 1.593037703762765e-05,
+      "loss": 0.6019,
+      "step": 4112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.222874533573527,
+      "learning_rate": 1.5928353711281138e-05,
+      "loss": 0.5621,
+      "step": 4113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1762218473594288,
+      "learning_rate": 1.5926330010647074e-05,
+      "loss": 0.6045,
+      "step": 4114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0863788174540303,
+      "learning_rate": 1.5924305935853218e-05,
+      "loss": 0.5308,
+      "step": 4115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2315550837648002,
+      "learning_rate": 1.5922281487027363e-05,
+      "loss": 0.6585,
+      "step": 4116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.156076057340976,
+      "learning_rate": 1.5920256664297326e-05,
+      "loss": 0.6323,
+      "step": 4117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0992176720153686,
+      "learning_rate": 1.5918231467790938e-05,
+      "loss": 0.576,
+      "step": 4118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.254389399439678,
+      "learning_rate": 1.5916205897636063e-05,
+      "loss": 0.6041,
+      "step": 4119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.159064758825171,
+      "learning_rate": 1.5914179953960584e-05,
+      "loss": 0.6353,
+      "step": 4120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.172107927697421,
+      "learning_rate": 1.5912153636892416e-05,
+      "loss": 0.5914,
+      "step": 4121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1033789085836685,
+      "learning_rate": 1.5910126946559484e-05,
+      "loss": 0.5793,
+      "step": 4122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1070690097439773,
+      "learning_rate": 1.5908099883089746e-05,
+      "loss": 0.5886,
+      "step": 4123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2302395077253894,
+      "learning_rate": 1.590607244661118e-05,
+      "loss": 0.5626,
+      "step": 4124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2806528607265886,
+      "learning_rate": 1.5904044637251793e-05,
+      "loss": 0.6355,
+      "step": 4125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1804876077333568,
+      "learning_rate": 1.5902016455139603e-05,
+      "loss": 0.6016,
+      "step": 4126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1975104873452846,
+      "learning_rate": 1.589998790040266e-05,
+      "loss": 0.5445,
+      "step": 4127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.251928082243929,
+      "learning_rate": 1.589795897316905e-05,
+      "loss": 0.572,
+      "step": 4128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.20924101094676,
+      "learning_rate": 1.5895929673566858e-05,
+      "loss": 0.6552,
+      "step": 4129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1965935019987255,
+      "learning_rate": 1.5893900001724204e-05,
+      "loss": 0.5841,
+      "step": 4130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2484053453642063,
+      "learning_rate": 1.5891869957769232e-05,
+      "loss": 0.6036,
+      "step": 4131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2915145210161616,
+      "learning_rate": 1.5889839541830115e-05,
+      "loss": 0.5962,
+      "step": 4132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.182488104455969,
+      "learning_rate": 1.5887808754035037e-05,
+      "loss": 0.6002,
+      "step": 4133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1066614194155773,
+      "learning_rate": 1.588577759451222e-05,
+      "loss": 0.5645,
+      "step": 4134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3685779585342646,
+      "learning_rate": 1.5883746063389897e-05,
+      "loss": 0.5852,
+      "step": 4135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.19871947988535,
+      "learning_rate": 1.5881714160796326e-05,
+      "loss": 0.6218,
+      "step": 4136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2420773727255203,
+      "learning_rate": 1.5879681886859794e-05,
+      "loss": 0.6229,
+      "step": 4137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2308886099309224,
+      "learning_rate": 1.5877649241708613e-05,
+      "loss": 0.5782,
+      "step": 4138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1947508845177341,
+      "learning_rate": 1.5875616225471105e-05,
+      "loss": 0.5959,
+      "step": 4139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1290108919319388,
+      "learning_rate": 1.5873582838275637e-05,
+      "loss": 0.5318,
+      "step": 4140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1401085925666197,
+      "learning_rate": 1.5871549080250577e-05,
+      "loss": 0.5707,
+      "step": 4141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1563511881497714,
+      "learning_rate": 1.586951495152433e-05,
+      "loss": 0.5712,
+      "step": 4142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.203907947180144,
+      "learning_rate": 1.5867480452225323e-05,
+      "loss": 0.6535,
+      "step": 4143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2168094150637931,
+      "learning_rate": 1.5865445582482002e-05,
+      "loss": 0.5632,
+      "step": 4144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2416731051250345,
+      "learning_rate": 1.586341034242284e-05,
+      "loss": 0.6188,
+      "step": 4145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2532720655488314,
+      "learning_rate": 1.5861374732176332e-05,
+      "loss": 0.6122,
+      "step": 4146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2004172831441842,
+      "learning_rate": 1.5859338751870998e-05,
+      "loss": 0.6108,
+      "step": 4147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0671746695823765,
+      "learning_rate": 1.5857302401635373e-05,
+      "loss": 0.5586,
+      "step": 4148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0990102433273106,
+      "learning_rate": 1.5855265681598032e-05,
+      "loss": 0.5914,
+      "step": 4149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.32162428478069,
+      "learning_rate": 1.585322859188756e-05,
+      "loss": 0.5657,
+      "step": 4150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1037112442418504,
+      "learning_rate": 1.5851191132632563e-05,
+      "loss": 0.591,
+      "step": 4151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2173754937429158,
+      "learning_rate": 1.584915330396169e-05,
+      "loss": 0.6426,
+      "step": 4152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.138650203649102,
+      "learning_rate": 1.5847115106003585e-05,
+      "loss": 0.5402,
+      "step": 4153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2264169497295956,
+      "learning_rate": 1.5845076538886934e-05,
+      "loss": 0.5583,
+      "step": 4154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2103888960961207,
+      "learning_rate": 1.584303760274045e-05,
+      "loss": 0.5968,
+      "step": 4155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.100631374446816,
+      "learning_rate": 1.5840998297692854e-05,
+      "loss": 0.5782,
+      "step": 4156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3140736637253614,
+      "learning_rate": 1.5838958623872902e-05,
+      "loss": 0.582,
+      "step": 4157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2300131302419608,
+      "learning_rate": 1.5836918581409365e-05,
+      "loss": 0.558,
+      "step": 4158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1115516478035843,
+      "learning_rate": 1.583487817043104e-05,
+      "loss": 0.5221,
+      "step": 4159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1952403925669344,
+      "learning_rate": 1.583283739106676e-05,
+      "loss": 0.6187,
+      "step": 4160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.197104911534478,
+      "learning_rate": 1.5830796243445357e-05,
+      "loss": 0.5644,
+      "step": 4161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0958404726924815,
+      "learning_rate": 1.5828754727695703e-05,
+      "loss": 0.5338,
+      "step": 4162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1056213273612223,
+      "learning_rate": 1.5826712843946693e-05,
+      "loss": 0.5319,
+      "step": 4163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1494895113001469,
+      "learning_rate": 1.582467059232724e-05,
+      "loss": 0.5539,
+      "step": 4164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1224026897365564,
+      "learning_rate": 1.582262797296628e-05,
+      "loss": 0.5406,
+      "step": 4165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.29688478948736,
+      "learning_rate": 1.5820584985992777e-05,
+      "loss": 0.5881,
+      "step": 4166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1768997803204808,
+      "learning_rate": 1.581854163153571e-05,
+      "loss": 0.5997,
+      "step": 4167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1589427214087622,
+      "learning_rate": 1.581649790972409e-05,
+      "loss": 0.5597,
+      "step": 4168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.134272456723087,
+      "learning_rate": 1.581445382068695e-05,
+      "loss": 0.561,
+      "step": 4169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1222380536812362,
+      "learning_rate": 1.5812409364553344e-05,
+      "loss": 0.6166,
+      "step": 4170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2356243329970122,
+      "learning_rate": 1.5810364541452342e-05,
+      "loss": 0.6328,
+      "step": 4171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1618939881314763,
+      "learning_rate": 1.580831935151305e-05,
+      "loss": 0.6192,
+      "step": 4172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1489546253816612,
+      "learning_rate": 1.5806273794864592e-05,
+      "loss": 0.5828,
+      "step": 4173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1931558738298869,
+      "learning_rate": 1.5804227871636114e-05,
+      "loss": 0.6067,
+      "step": 4174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.224700340287395,
+      "learning_rate": 1.5802181581956782e-05,
+      "loss": 0.6024,
+      "step": 4175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.197164658836633,
+      "learning_rate": 1.5800134925955792e-05,
+      "loss": 0.6414,
+      "step": 4176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0887383217957185,
+      "learning_rate": 1.579808790376236e-05,
+      "loss": 0.6058,
+      "step": 4177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.259549996307871,
+      "learning_rate": 1.5796040515505724e-05,
+      "loss": 0.6651,
+      "step": 4178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0770380373394608,
+      "learning_rate": 1.5793992761315147e-05,
+      "loss": 0.5502,
+      "step": 4179
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2953728398348576,
+      "learning_rate": 1.5791944641319914e-05,
+      "loss": 0.6397,
+      "step": 4180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2637665372647529,
+      "learning_rate": 1.5789896155649333e-05,
+      "loss": 0.5922,
+      "step": 4181
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2070424594790734,
+      "learning_rate": 1.578784730443273e-05,
+      "loss": 0.5758,
+      "step": 4182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.290383557880392,
+      "learning_rate": 1.5785798087799476e-05,
+      "loss": 0.5818,
+      "step": 4183
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2466818160311268,
+      "learning_rate": 1.5783748505878932e-05,
+      "loss": 0.6033,
+      "step": 4184
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.371210468247492,
+      "learning_rate": 1.5781698558800503e-05,
+      "loss": 0.6262,
+      "step": 4185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.185993694838671,
+      "learning_rate": 1.577964824669362e-05,
+      "loss": 0.6076,
+      "step": 4186
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.190057510140332,
+      "learning_rate": 1.577759756968772e-05,
+      "loss": 0.5602,
+      "step": 4187
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1809004554076394,
+      "learning_rate": 1.577554652791228e-05,
+      "loss": 0.5978,
+      "step": 4188
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.238725889261915,
+      "learning_rate": 1.5773495121496787e-05,
+      "loss": 0.5387,
+      "step": 4189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.12641822642419,
+      "learning_rate": 1.5771443350570756e-05,
+      "loss": 0.6105,
+      "step": 4190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2080687474368461,
+      "learning_rate": 1.5769391215263737e-05,
+      "loss": 0.5477,
+      "step": 4191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.189125554622287,
+      "learning_rate": 1.576733871570528e-05,
+      "loss": 0.6163,
+      "step": 4192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1753070166413515,
+      "learning_rate": 1.576528585202498e-05,
+      "loss": 0.6237,
+      "step": 4193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2476602591281176,
+      "learning_rate": 1.576323262435243e-05,
+      "loss": 0.5966,
+      "step": 4194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1575786071187284,
+      "learning_rate": 1.5761179032817275e-05,
+      "loss": 0.56,
+      "step": 4195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.247796787284363,
+      "learning_rate": 1.5759125077549164e-05,
+      "loss": 0.5504,
+      "step": 4196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2886518344436713,
+      "learning_rate": 1.5757070758677775e-05,
+      "loss": 0.6028,
+      "step": 4197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1983164521002307,
+      "learning_rate": 1.5755016076332805e-05,
+      "loss": 0.5647,
+      "step": 4198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1704388783240482,
+      "learning_rate": 1.5752961030643978e-05,
+      "loss": 0.5942,
+      "step": 4199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1912501999013279,
+      "learning_rate": 1.5750905621741037e-05,
+      "loss": 0.6125,
+      "step": 4200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.135771156525432,
+      "learning_rate": 1.5748849849753757e-05,
+      "loss": 0.5413,
+      "step": 4201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.084404226563342,
+      "learning_rate": 1.5746793714811925e-05,
+      "loss": 0.5594,
+      "step": 4202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.177701124035044,
+      "learning_rate": 1.5744737217045355e-05,
+      "loss": 0.5573,
+      "step": 4203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1667078669403046,
+      "learning_rate": 1.574268035658389e-05,
+      "loss": 0.5941,
+      "step": 4204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3080694714100287,
+      "learning_rate": 1.574062313355738e-05,
+      "loss": 0.6122,
+      "step": 4205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1308704434974841,
+      "learning_rate": 1.5738565548095718e-05,
+      "loss": 0.5881,
+      "step": 4206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2058847160877186,
+      "learning_rate": 1.5736507600328804e-05,
+      "loss": 0.6081,
+      "step": 4207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3394470072306475,
+      "learning_rate": 1.5734449290386568e-05,
+      "loss": 0.6675,
+      "step": 4208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3144116331356501,
+      "learning_rate": 1.5732390618398966e-05,
+      "loss": 0.6165,
+      "step": 4209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1398986698591642,
+      "learning_rate": 1.5730331584495965e-05,
+      "loss": 0.6137,
+      "step": 4210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1406860335233333,
+      "learning_rate": 1.5728272188807564e-05,
+      "loss": 0.5765,
+      "step": 4211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1929319025031162,
+      "learning_rate": 1.572621243146379e-05,
+      "loss": 0.5761,
+      "step": 4212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2316240004056611,
+      "learning_rate": 1.5724152312594683e-05,
+      "loss": 0.5776,
+      "step": 4213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2885305983433712,
+      "learning_rate": 1.5722091832330307e-05,
+      "loss": 0.6087,
+      "step": 4214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2260871014512607,
+      "learning_rate": 1.572003099080075e-05,
+      "loss": 0.5777,
+      "step": 4215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0597741559392169,
+      "learning_rate": 1.5717969788136123e-05,
+      "loss": 0.5589,
+      "step": 4216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2328517544954958,
+      "learning_rate": 1.571590822446657e-05,
+      "loss": 0.5962,
+      "step": 4217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.180643818242358,
+      "learning_rate": 1.571384629992223e-05,
+      "loss": 0.6047,
+      "step": 4218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.153029208869553,
+      "learning_rate": 1.57117840146333e-05,
+      "loss": 0.6125,
+      "step": 4219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.122415912681638,
+      "learning_rate": 1.5709721368729977e-05,
+      "loss": 0.5619,
+      "step": 4220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2249023963035894,
+      "learning_rate": 1.5707658362342484e-05,
+      "loss": 0.6708,
+      "step": 4221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.343392124869427,
+      "learning_rate": 1.570559499560107e-05,
+      "loss": 0.5913,
+      "step": 4222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2005444880849245,
+      "learning_rate": 1.570353126863601e-05,
+      "loss": 0.6078,
+      "step": 4223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1056571233814998,
+      "learning_rate": 1.570146718157759e-05,
+      "loss": 0.5583,
+      "step": 4224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1242107165740816,
+      "learning_rate": 1.5699402734556133e-05,
+      "loss": 0.5937,
+      "step": 4225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1543452580311813,
+      "learning_rate": 1.5697337927701977e-05,
+      "loss": 0.5595,
+      "step": 4226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.267246008861621,
+      "learning_rate": 1.5695272761145486e-05,
+      "loss": 0.6382,
+      "step": 4227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1132376143620815,
+      "learning_rate": 1.569320723501704e-05,
+      "loss": 0.587,
+      "step": 4228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.164809838357185,
+      "learning_rate": 1.5691141349447046e-05,
+      "loss": 0.5429,
+      "step": 4229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2153910296516017,
+      "learning_rate": 1.5689075104565936e-05,
+      "loss": 0.6058,
+      "step": 4230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1422446783237805,
+      "learning_rate": 1.5687008500504165e-05,
+      "loss": 0.6221,
+      "step": 4231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1590503597986488,
+      "learning_rate": 1.5684941537392205e-05,
+      "loss": 0.6127,
+      "step": 4232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2456341795863013,
+      "learning_rate": 1.5682874215360557e-05,
+      "loss": 0.5934,
+      "step": 4233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2216283115040354,
+      "learning_rate": 1.568080653453974e-05,
+      "loss": 0.5776,
+      "step": 4234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0893601195587912,
+      "learning_rate": 1.5678738495060292e-05,
+      "loss": 0.5819,
+      "step": 4235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0774837743152603,
+      "learning_rate": 1.567667009705279e-05,
+      "loss": 0.509,
+      "step": 4236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0842760947906658,
+      "learning_rate": 1.567460134064782e-05,
+      "loss": 0.5292,
+      "step": 4237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1248018832071582,
+      "learning_rate": 1.5672532225975983e-05,
+      "loss": 0.5691,
+      "step": 4238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.189464800797708,
+      "learning_rate": 1.5670462753167922e-05,
+      "loss": 0.5998,
+      "step": 4239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3506471036684864,
+      "learning_rate": 1.566839292235429e-05,
+      "loss": 0.6264,
+      "step": 4240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1231247319309805,
+      "learning_rate": 1.5666322733665773e-05,
+      "loss": 0.6178,
+      "step": 4241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1648143414086092,
+      "learning_rate": 1.5664252187233066e-05,
+      "loss": 0.5785,
+      "step": 4242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2548386859185994,
+      "learning_rate": 1.5662181283186894e-05,
+      "loss": 0.6065,
+      "step": 4243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2017569752874238,
+      "learning_rate": 1.5660110021658002e-05,
+      "loss": 0.5535,
+      "step": 4244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1170711323347944,
+      "learning_rate": 1.5658038402777165e-05,
+      "loss": 0.5489,
+      "step": 4245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1981973678551041,
+      "learning_rate": 1.5655966426675172e-05,
+      "loss": 0.6382,
+      "step": 4246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2080716090830148,
+      "learning_rate": 1.5653894093482835e-05,
+      "loss": 0.5793,
+      "step": 4247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1014751102061386,
+      "learning_rate": 1.5651821403330996e-05,
+      "loss": 0.5579,
+      "step": 4248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1856763288279222,
+      "learning_rate": 1.5649748356350513e-05,
+      "loss": 0.6111,
+      "step": 4249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.111876608818996,
+      "learning_rate": 1.5647674952672265e-05,
+      "loss": 0.5396,
+      "step": 4250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1950902794609333,
+      "learning_rate": 1.564560119242716e-05,
+      "loss": 0.681,
+      "step": 4251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2280012306182362,
+      "learning_rate": 1.5643527075746125e-05,
+      "loss": 0.5947,
+      "step": 4252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1791713166341768,
+      "learning_rate": 1.564145260276011e-05,
+      "loss": 0.5775,
+      "step": 4253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2863204460019095,
+      "learning_rate": 1.563937777360008e-05,
+      "loss": 0.6008,
+      "step": 4254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.15050471052642,
+      "learning_rate": 1.563730258839704e-05,
+      "loss": 0.6132,
+      "step": 4255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1993216305816103,
+      "learning_rate": 1.5635227047282005e-05,
+      "loss": 0.5771,
+      "step": 4256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1615289344484334,
+      "learning_rate": 1.563315115038601e-05,
+      "loss": 0.561,
+      "step": 4257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1735187954431585,
+      "learning_rate": 1.563107489784012e-05,
+      "loss": 0.5682,
+      "step": 4258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2188851697253185,
+      "learning_rate": 1.562899828977542e-05,
+      "loss": 0.5923,
+      "step": 4259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1471939100867332,
+      "learning_rate": 1.5626921326323016e-05,
+      "loss": 0.5777,
+      "step": 4260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1356480330638674,
+      "learning_rate": 1.5624844007614037e-05,
+      "loss": 0.6542,
+      "step": 4261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.295458192074889,
+      "learning_rate": 1.5622766333779637e-05,
+      "loss": 0.6568,
+      "step": 4262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1669861095062695,
+      "learning_rate": 1.5620688304950985e-05,
+      "loss": 0.6234,
+      "step": 4263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.186047317989056,
+      "learning_rate": 1.5618609921259286e-05,
+      "loss": 0.5964,
+      "step": 4264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.238075503997241,
+      "learning_rate": 1.561653118283575e-05,
+      "loss": 0.5791,
+      "step": 4265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1897992912898259,
+      "learning_rate": 1.5614452089811628e-05,
+      "loss": 0.6058,
+      "step": 4266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0708700245509777,
+      "learning_rate": 1.5612372642318176e-05,
+      "loss": 0.5759,
+      "step": 4267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1604674978212233,
+      "learning_rate": 1.5610292840486684e-05,
+      "loss": 0.5637,
+      "step": 4268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1496032714681725,
+      "learning_rate": 1.5608212684448453e-05,
+      "loss": 0.6144,
+      "step": 4269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2419530299361183,
+      "learning_rate": 1.560613217433483e-05,
+      "loss": 0.6095,
+      "step": 4270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2836035534791388,
+      "learning_rate": 1.5604051310277152e-05,
+      "loss": 0.6439,
+      "step": 4271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1591084176371425,
+      "learning_rate": 1.5601970092406807e-05,
+      "loss": 0.5686,
+      "step": 4272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.13279150581637,
+      "learning_rate": 1.5599888520855186e-05,
+      "loss": 0.5598,
+      "step": 4273
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.203480606040116,
+      "learning_rate": 1.559780659575371e-05,
+      "loss": 0.588,
+      "step": 4274
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.130988816938756,
+      "learning_rate": 1.559572431723382e-05,
+      "loss": 0.5794,
+      "step": 4275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2944949728027186,
+      "learning_rate": 1.559364168542699e-05,
+      "loss": 0.6655,
+      "step": 4276
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1272260047928933,
+      "learning_rate": 1.55915587004647e-05,
+      "loss": 0.5523,
+      "step": 4277
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2182135746373812,
+      "learning_rate": 1.5589475362478458e-05,
+      "loss": 0.5855,
+      "step": 4278
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3181940271744912,
+      "learning_rate": 1.5587391671599798e-05,
+      "loss": 0.6113,
+      "step": 4279
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2505427612686963,
+      "learning_rate": 1.5585307627960277e-05,
+      "loss": 0.5963,
+      "step": 4280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0647500840075304,
+      "learning_rate": 1.5583223231691466e-05,
+      "loss": 0.5443,
+      "step": 4281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0774165604007833,
+      "learning_rate": 1.558113848292497e-05,
+      "loss": 0.5454,
+      "step": 4282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.109701994938504,
+      "learning_rate": 1.5579053381792406e-05,
+      "loss": 0.5765,
+      "step": 4283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1624594794153704,
+      "learning_rate": 1.5576967928425414e-05,
+      "loss": 0.5757,
+      "step": 4284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.281872388817533,
+      "learning_rate": 1.5574882122955668e-05,
+      "loss": 0.5887,
+      "step": 4285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1957303482441077,
+      "learning_rate": 1.557279596551485e-05,
+      "loss": 0.5552,
+      "step": 4286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1569548855346417,
+      "learning_rate": 1.557070945623467e-05,
+      "loss": 0.6048,
+      "step": 4287
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1554446121800024,
+      "learning_rate": 1.556862259524686e-05,
+      "loss": 0.5987,
+      "step": 4288
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0879999239409643,
+      "learning_rate": 1.556653538268318e-05,
+      "loss": 0.5325,
+      "step": 4289
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2413542249856198,
+      "learning_rate": 1.5564447818675397e-05,
+      "loss": 0.5897,
+      "step": 4290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1876448492750868,
+      "learning_rate": 1.5562359903355315e-05,
+      "loss": 0.5331,
+      "step": 4291
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1003812584361574,
+      "learning_rate": 1.5560271636854757e-05,
+      "loss": 0.5632,
+      "step": 4292
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1683119457824958,
+      "learning_rate": 1.555818301930556e-05,
+      "loss": 0.5694,
+      "step": 4293
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.112776524755778,
+      "learning_rate": 1.555609405083959e-05,
+      "loss": 0.5432,
+      "step": 4294
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3055725293825904,
+      "learning_rate": 1.5554004731588745e-05,
+      "loss": 0.6332,
+      "step": 4295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1394203827357794,
+      "learning_rate": 1.555191506168492e-05,
+      "loss": 0.5719,
+      "step": 4296
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3082075317709538,
+      "learning_rate": 1.5549825041260052e-05,
+      "loss": 0.624,
+      "step": 4297
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.341341388358419,
+      "learning_rate": 1.5547734670446103e-05,
+      "loss": 0.6568,
+      "step": 4298
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2121439184785832,
+      "learning_rate": 1.554564394937504e-05,
+      "loss": 0.6046,
+      "step": 4299
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3055558199014239,
+      "learning_rate": 1.5543552878178857e-05,
+      "loss": 0.6063,
+      "step": 4300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2305411181285864,
+      "learning_rate": 1.5541461456989583e-05,
+      "loss": 0.649,
+      "step": 4301
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.192278983229602,
+      "learning_rate": 1.553936968593926e-05,
+      "loss": 0.6031,
+      "step": 4302
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0950641366822178,
+      "learning_rate": 1.5537277565159944e-05,
+      "loss": 0.6649,
+      "step": 4303
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1334477715122804,
+      "learning_rate": 1.5535185094783728e-05,
+      "loss": 0.5492,
+      "step": 4304
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1961000129640387,
+      "learning_rate": 1.5533092274942724e-05,
+      "loss": 0.6009,
+      "step": 4305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.266119118767877,
+      "learning_rate": 1.553099910576905e-05,
+      "loss": 0.5972,
+      "step": 4306
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2080482223277254,
+      "learning_rate": 1.5528905587394872e-05,
+      "loss": 0.5804,
+      "step": 4307
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.205405958892666,
+      "learning_rate": 1.5526811719952356e-05,
+      "loss": 0.6101,
+      "step": 4308
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1561293667935484,
+      "learning_rate": 1.55247175035737e-05,
+      "loss": 0.6169,
+      "step": 4309
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2087806443725497,
+      "learning_rate": 1.5522622938391132e-05,
+      "loss": 0.6186,
+      "step": 4310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2466818160311268,
+      "learning_rate": 1.552052802453688e-05,
+      "loss": 0.6416,
+      "step": 4311
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2414732024933746,
+      "learning_rate": 1.551843276214321e-05,
+      "loss": 0.5562,
+      "step": 4312
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3189507860397496,
+      "learning_rate": 1.551633715134241e-05,
+      "loss": 0.669,
+      "step": 4313
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2859813047714257,
+      "learning_rate": 1.5514241192266786e-05,
+      "loss": 0.6141,
+      "step": 4314
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0726381637926647,
+      "learning_rate": 1.5512144885048664e-05,
+      "loss": 0.6105,
+      "step": 4315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2531460748360561,
+      "learning_rate": 1.5510048229820398e-05,
+      "loss": 0.5823,
+      "step": 4316
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2201612078360424,
+      "learning_rate": 1.5507951226714356e-05,
+      "loss": 0.5963,
+      "step": 4317
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2030143686928403,
+      "learning_rate": 1.550585387586294e-05,
+      "loss": 0.6031,
+      "step": 4318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1650962591720035,
+      "learning_rate": 1.550375617739856e-05,
+      "loss": 0.6219,
+      "step": 4319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2110246196294916,
+      "learning_rate": 1.550165813145366e-05,
+      "loss": 0.5959,
+      "step": 4320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2279511869868345,
+      "learning_rate": 1.5499559738160693e-05,
+      "loss": 0.6043,
+      "step": 4321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.063445568070926,
+      "learning_rate": 1.549746099765215e-05,
+      "loss": 0.5189,
+      "step": 4322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1293839232741927,
+      "learning_rate": 1.5495361910060527e-05,
+      "loss": 0.5382,
+      "step": 4323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1155763759755608,
+      "learning_rate": 1.5493262475518353e-05,
+      "loss": 0.6106,
+      "step": 4324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1945891341960149,
+      "learning_rate": 1.5491162694158182e-05,
+      "loss": 0.5521,
+      "step": 4325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3651839480275945,
+      "learning_rate": 1.548906256611258e-05,
+      "loss": 0.609,
+      "step": 4326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2565710445133467,
+      "learning_rate": 1.5486962091514133e-05,
+      "loss": 0.6111,
+      "step": 4327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1392565840530247,
+      "learning_rate": 1.5484861270495464e-05,
+      "loss": 0.5218,
+      "step": 4328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2287283078022297,
+      "learning_rate": 1.5482760103189203e-05,
+      "loss": 0.6049,
+      "step": 4329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2041847713829785,
+      "learning_rate": 1.548065858972801e-05,
+      "loss": 0.5831,
+      "step": 4330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1843491966637465,
+      "learning_rate": 1.5478556730244564e-05,
+      "loss": 0.5631,
+      "step": 4331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0813510560897177,
+      "learning_rate": 1.5476454524871566e-05,
+      "loss": 0.5709,
+      "step": 4332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.194321614364365,
+      "learning_rate": 1.5474351973741742e-05,
+      "loss": 0.6366,
+      "step": 4333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2448822158801993,
+      "learning_rate": 1.547224907698783e-05,
+      "loss": 0.6267,
+      "step": 4334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1498157768304422,
+      "learning_rate": 1.54701458347426e-05,
+      "loss": 0.5494,
+      "step": 4335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0574603329877628,
+      "learning_rate": 1.5468042247138844e-05,
+      "loss": 0.5413,
+      "step": 4336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1500462273967669,
+      "learning_rate": 1.5465938314309367e-05,
+      "loss": 0.5746,
+      "step": 4337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1411223176558392,
+      "learning_rate": 1.546383403638701e-05,
+      "loss": 0.5566,
+      "step": 4338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.225857137029262,
+      "learning_rate": 1.5461729413504613e-05,
+      "loss": 0.6297,
+      "step": 4339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0602756666311752,
+      "learning_rate": 1.5459624445795062e-05,
+      "loss": 0.5592,
+      "step": 4340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.113962172689552,
+      "learning_rate": 1.545751913339125e-05,
+      "loss": 0.5249,
+      "step": 4341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2147959610840169,
+      "learning_rate": 1.54554134764261e-05,
+      "loss": 0.6144,
+      "step": 4342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.189400657661709,
+      "learning_rate": 1.5453307475032552e-05,
+      "loss": 0.5544,
+      "step": 4343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1880256091931498,
+      "learning_rate": 1.5451201129343566e-05,
+      "loss": 0.5899,
+      "step": 4344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1813865703774582,
+      "learning_rate": 1.544909443949213e-05,
+      "loss": 0.6005,
+      "step": 4345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0560182441305859,
+      "learning_rate": 1.5446987405611248e-05,
+      "loss": 0.5167,
+      "step": 4346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1708212119908372,
+      "learning_rate": 1.5444880027833947e-05,
+      "loss": 0.5678,
+      "step": 4347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2678871189618748,
+      "learning_rate": 1.5442772306293277e-05,
+      "loss": 0.6456,
+      "step": 4348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.081308006097804,
+      "learning_rate": 1.5440664241122312e-05,
+      "loss": 0.5323,
+      "step": 4349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2158441380383915,
+      "learning_rate": 1.5438555832454143e-05,
+      "loss": 0.6012,
+      "step": 4350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1261988715655107,
+      "learning_rate": 1.5436447080421887e-05,
+      "loss": 0.6277,
+      "step": 4351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1096401166454524,
+      "learning_rate": 1.5434337985158674e-05,
+      "loss": 0.5559,
+      "step": 4352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2222174201254479,
+      "learning_rate": 1.5432228546797668e-05,
+      "loss": 0.5718,
+      "step": 4353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2757120238640933,
+      "learning_rate": 1.5430118765472052e-05,
+      "loss": 0.582,
+      "step": 4354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1499292973884567,
+      "learning_rate": 1.5428008641315018e-05,
+      "loss": 0.5528,
+      "step": 4355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1802928457727033,
+      "learning_rate": 1.5425898174459794e-05,
+      "loss": 0.5814,
+      "step": 4356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0787393782038581,
+      "learning_rate": 1.5423787365039627e-05,
+      "loss": 0.533,
+      "step": 4357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2340041460588937,
+      "learning_rate": 1.5421676213187774e-05,
+      "loss": 0.6392,
+      "step": 4358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2280317606002826,
+      "learning_rate": 1.5419564719037536e-05,
+      "loss": 0.6096,
+      "step": 4359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.248507371458292,
+      "learning_rate": 1.5417452882722214e-05,
+      "loss": 0.6251,
+      "step": 4360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.226481441079024,
+      "learning_rate": 1.541534070437514e-05,
+      "loss": 0.6905,
+      "step": 4361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1224117174669375,
+      "learning_rate": 1.541322818412967e-05,
+      "loss": 0.5736,
+      "step": 4362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1324407033154424,
+      "learning_rate": 1.5411115322119176e-05,
+      "loss": 0.5572,
+      "step": 4363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.25963105658449,
+      "learning_rate": 1.5409002118477053e-05,
+      "loss": 0.6141,
+      "step": 4364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2681653465594676,
+      "learning_rate": 1.540688857333672e-05,
+      "loss": 0.599,
+      "step": 4365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.214661121440196,
+      "learning_rate": 1.5404774686831615e-05,
+      "loss": 0.6299,
+      "step": 4366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.21888746806788,
+      "learning_rate": 1.54026604590952e-05,
+      "loss": 0.5957,
+      "step": 4367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1919438884412488,
+      "learning_rate": 1.540054589026095e-05,
+      "loss": 0.5919,
+      "step": 4368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2466692895827691,
+      "learning_rate": 1.5398430980462382e-05,
+      "loss": 0.5912,
+      "step": 4369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1907434339516478,
+      "learning_rate": 1.5396315729833015e-05,
+      "loss": 0.5307,
+      "step": 4370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1197120046742197,
+      "learning_rate": 1.5394200138506393e-05,
+      "loss": 0.5414,
+      "step": 4371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1677851708092752,
+      "learning_rate": 1.5392084206616084e-05,
+      "loss": 0.5809,
+      "step": 4372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1437043206937372,
+      "learning_rate": 1.5389967934295677e-05,
+      "loss": 0.5991,
+      "step": 4373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0779447543084313,
+      "learning_rate": 1.5387851321678788e-05,
+      "loss": 0.5676,
+      "step": 4374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2527462355953578,
+      "learning_rate": 1.538573436889905e-05,
+      "loss": 0.6305,
+      "step": 4375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1712196043257672,
+      "learning_rate": 1.5383617076090114e-05,
+      "loss": 0.6145,
+      "step": 4376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1639391430516126,
+      "learning_rate": 1.5381499443385653e-05,
+      "loss": 0.5949,
+      "step": 4377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1495743917624666,
+      "learning_rate": 1.537938147091937e-05,
+      "loss": 0.5887,
+      "step": 4378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3407043273504264,
+      "learning_rate": 1.537726315882498e-05,
+      "loss": 0.6772,
+      "step": 4379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.250110859723331,
+      "learning_rate": 1.5375144507236222e-05,
+      "loss": 0.5646,
+      "step": 4380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2708810526325784,
+      "learning_rate": 1.537302551628686e-05,
+      "loss": 0.6346,
+      "step": 4381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0891656988169294,
+      "learning_rate": 1.5370906186110677e-05,
+      "loss": 0.5442,
+      "step": 4382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2144545149281407,
+      "learning_rate": 1.536878651684148e-05,
+      "loss": 0.5854,
+      "step": 4383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2673010853991138,
+      "learning_rate": 1.5366666508613083e-05,
+      "loss": 0.6416,
+      "step": 4384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1962630537670147,
+      "learning_rate": 1.536454616155935e-05,
+      "loss": 0.5655,
+      "step": 4385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2514959443834761,
+      "learning_rate": 1.5362425475814133e-05,
+      "loss": 0.5454,
+      "step": 4386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1835336701406634,
+      "learning_rate": 1.5360304451511333e-05,
+      "loss": 0.5463,
+      "step": 4387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2501789441770663,
+      "learning_rate": 1.5358183088784853e-05,
+      "loss": 0.5776,
+      "step": 4388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2462837767926083,
+      "learning_rate": 1.5356061387768634e-05,
+      "loss": 0.5915,
+      "step": 4389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1739182524903378,
+      "learning_rate": 1.535393934859663e-05,
+      "loss": 0.524,
+      "step": 4390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.230764153761204,
+      "learning_rate": 1.5351816971402803e-05,
+      "loss": 0.6316,
+      "step": 4391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.097198119213983,
+      "learning_rate": 1.5349694256321162e-05,
+      "loss": 0.5255,
+      "step": 4392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0370746613923232,
+      "learning_rate": 1.5347571203485723e-05,
+      "loss": 0.5667,
+      "step": 4393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0880541584370695,
+      "learning_rate": 1.5345447813030526e-05,
+      "loss": 0.5446,
+      "step": 4394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2309443932035438,
+      "learning_rate": 1.5343324085089628e-05,
+      "loss": 0.6333,
+      "step": 4395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1745729542887466,
+      "learning_rate": 1.534120001979711e-05,
+      "loss": 0.5948,
+      "step": 4396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2211402537642444,
+      "learning_rate": 1.533907561728708e-05,
+      "loss": 0.5645,
+      "step": 4397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1916706729812678,
+      "learning_rate": 1.533695087769366e-05,
+      "loss": 0.5642,
+      "step": 4398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2430540217432164,
+      "learning_rate": 1.5334825801150998e-05,
+      "loss": 0.6135,
+      "step": 4399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2546036346644387,
+      "learning_rate": 1.5332700387793255e-05,
+      "loss": 0.6196,
+      "step": 4400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2020119311427646,
+      "learning_rate": 1.5330574637754627e-05,
+      "loss": 0.5501,
+      "step": 4401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.234787932269557,
+      "learning_rate": 1.5328448551169318e-05,
+      "loss": 0.6314,
+      "step": 4402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.063317657452814,
+      "learning_rate": 1.532632212817156e-05,
+      "loss": 0.5592,
+      "step": 4403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1959654078281263,
+      "learning_rate": 1.532419536889561e-05,
+      "loss": 0.5993,
+      "step": 4404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2815247683027342,
+      "learning_rate": 1.5322068273475737e-05,
+      "loss": 0.5572,
+      "step": 4405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2080824142031188,
+      "learning_rate": 1.531994084204623e-05,
+      "loss": 0.5759,
+      "step": 4406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.131741036755641,
+      "learning_rate": 1.5317813074741415e-05,
+      "loss": 0.5293,
+      "step": 4407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2769330965737027,
+      "learning_rate": 1.531568497169562e-05,
+      "loss": 0.5712,
+      "step": 4408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2140100200214425,
+      "learning_rate": 1.5313556533043212e-05,
+      "loss": 0.6009,
+      "step": 4409
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2692971351141003,
+      "learning_rate": 1.5311427758918564e-05,
+      "loss": 0.6094,
+      "step": 4410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1314977448357104,
+      "learning_rate": 1.5309298649456075e-05,
+      "loss": 0.5783,
+      "step": 4411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.164858705768735,
+      "learning_rate": 1.5307169204790174e-05,
+      "loss": 0.6357,
+      "step": 4412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1666847238959757,
+      "learning_rate": 1.5305039425055302e-05,
+      "loss": 0.5311,
+      "step": 4413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1897781003297598,
+      "learning_rate": 1.5302909310385916e-05,
+      "loss": 0.5929,
+      "step": 4414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1927447195485912,
+      "learning_rate": 1.530077886091651e-05,
+      "loss": 0.5851,
+      "step": 4415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1762167292074122,
+      "learning_rate": 1.5298648076781583e-05,
+      "loss": 0.6102,
+      "step": 4416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0890326541001754,
+      "learning_rate": 1.5296516958115666e-05,
+      "loss": 0.5465,
+      "step": 4417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2519195123644502,
+      "learning_rate": 1.5294385505053305e-05,
+      "loss": 0.5832,
+      "step": 4418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.232027646714422,
+      "learning_rate": 1.5292253717729072e-05,
+      "loss": 0.6447,
+      "step": 4419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2657436268160225,
+      "learning_rate": 1.529012159627756e-05,
+      "loss": 0.6693,
+      "step": 4420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.210744042478969,
+      "learning_rate": 1.5287989140833376e-05,
+      "loss": 0.5477,
+      "step": 4421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0917837725622568,
+      "learning_rate": 1.5285856351531157e-05,
+      "loss": 0.58,
+      "step": 4422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.148882046769742,
+      "learning_rate": 1.5283723228505552e-05,
+      "loss": 0.5378,
+      "step": 4423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.083430518168269,
+      "learning_rate": 1.5281589771891244e-05,
+      "loss": 0.4977,
+      "step": 4424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1574323125208323,
+      "learning_rate": 1.527945598182292e-05,
+      "loss": 0.5491,
+      "step": 4425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1496577105923107,
+      "learning_rate": 1.5277321858435303e-05,
+      "loss": 0.5835,
+      "step": 4426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2745523545806017,
+      "learning_rate": 1.527518740186313e-05,
+      "loss": 0.6344,
+      "step": 4427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2802791638977342,
+      "learning_rate": 1.527305261224116e-05,
+      "loss": 0.615,
+      "step": 4428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.148809671107842,
+      "learning_rate": 1.5270917489704173e-05,
+      "loss": 0.5432,
+      "step": 4429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1369891287553175,
+      "learning_rate": 1.5268782034386972e-05,
+      "loss": 0.6384,
+      "step": 4430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0942323438559076,
+      "learning_rate": 1.5266646246424374e-05,
+      "loss": 0.5141,
+      "step": 4431
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1011347921080916,
+      "learning_rate": 1.5264510125951228e-05,
+      "loss": 0.563,
+      "step": 4432
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.088406178683373,
+      "learning_rate": 1.5262373673102396e-05,
+      "loss": 0.5599,
+      "step": 4433
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.302010536702029,
+      "learning_rate": 1.5260236888012766e-05,
+      "loss": 0.6163,
+      "step": 4434
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1073075492822713,
+      "learning_rate": 1.5258099770817242e-05,
+      "loss": 0.5581,
+      "step": 4435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.146683579771115,
+      "learning_rate": 1.525596232165075e-05,
+      "loss": 0.5596,
+      "step": 4436
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.170456752876025,
+      "learning_rate": 1.5253824540648237e-05,
+      "loss": 0.5416,
+      "step": 4437
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1076545807229496,
+      "learning_rate": 1.5251686427944679e-05,
+      "loss": 0.5367,
+      "step": 4438
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1734350882554954,
+      "learning_rate": 1.524954798367506e-05,
+      "loss": 0.563,
+      "step": 4439
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1514843856011263,
+      "learning_rate": 1.5247409207974394e-05,
+      "loss": 0.588,
+      "step": 4440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2255016525892417,
+      "learning_rate": 1.5245270100977707e-05,
+      "loss": 0.578,
+      "step": 4441
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.185576336414464,
+      "learning_rate": 1.5243130662820058e-05,
+      "loss": 0.5966,
+      "step": 4442
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0944559816106305,
+      "learning_rate": 1.5240990893636522e-05,
+      "loss": 0.6105,
+      "step": 4443
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1557597074503254,
+      "learning_rate": 1.523885079356219e-05,
+      "loss": 0.6533,
+      "step": 4444
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1560541451026403,
+      "learning_rate": 1.5236710362732178e-05,
+      "loss": 0.5969,
+      "step": 4445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0838196225445664,
+      "learning_rate": 1.5234569601281623e-05,
+      "loss": 0.5379,
+      "step": 4446
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2456933218234425,
+      "learning_rate": 1.523242850934568e-05,
+      "loss": 0.5772,
+      "step": 4447
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1409657113174614,
+      "learning_rate": 1.5230287087059532e-05,
+      "loss": 0.5509,
+      "step": 4448
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.100200054010176,
+      "learning_rate": 1.5228145334558377e-05,
+      "loss": 0.5919,
+      "step": 4449
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2097921553927675,
+      "learning_rate": 1.5226003251977432e-05,
+      "loss": 0.6066,
+      "step": 4450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2463753122060177,
+      "learning_rate": 1.5223860839451935e-05,
+      "loss": 0.5836,
+      "step": 4451
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.081399065006784,
+      "learning_rate": 1.5221718097117157e-05,
+      "loss": 0.5625,
+      "step": 4452
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.176278246906274,
+      "learning_rate": 1.5219575025108373e-05,
+      "loss": 0.5509,
+      "step": 4453
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.153852089842175,
+      "learning_rate": 1.521743162356089e-05,
+      "loss": 0.5421,
+      "step": 4454
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.238385457849631,
+      "learning_rate": 1.521528789261003e-05,
+      "loss": 0.6216,
+      "step": 4455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.128811472573839,
+      "learning_rate": 1.5213143832391133e-05,
+      "loss": 0.5805,
+      "step": 4456
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2818857337059608,
+      "learning_rate": 1.5210999443039573e-05,
+      "loss": 0.5872,
+      "step": 4457
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3346661323796543,
+      "learning_rate": 1.5208854724690734e-05,
+      "loss": 0.6064,
+      "step": 4458
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.055467105528188,
+      "learning_rate": 1.5206709677480022e-05,
+      "loss": 0.5526,
+      "step": 4459
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1356789988736544,
+      "learning_rate": 1.5204564301542863e-05,
+      "loss": 0.5718,
+      "step": 4460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.159306070175738,
+      "learning_rate": 1.520241859701471e-05,
+      "loss": 0.6041,
+      "step": 4461
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1956427124274813,
+      "learning_rate": 1.5200272564031026e-05,
+      "loss": 0.5949,
+      "step": 4462
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2159479648725056,
+      "learning_rate": 1.5198126202727311e-05,
+      "loss": 0.6024,
+      "step": 4463
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1560298091417576,
+      "learning_rate": 1.5195979513239064e-05,
+      "loss": 0.5788,
+      "step": 4464
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1778144872440648,
+      "learning_rate": 1.5193832495701825e-05,
+      "loss": 0.5278,
+      "step": 4465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.168228732871617,
+      "learning_rate": 1.5191685150251146e-05,
+      "loss": 0.586,
+      "step": 4466
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2033463806984266,
+      "learning_rate": 1.5189537477022595e-05,
+      "loss": 0.6291,
+      "step": 4467
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0862332497063598,
+      "learning_rate": 1.518738947615177e-05,
+      "loss": 0.5285,
+      "step": 4468
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1921114976087672,
+      "learning_rate": 1.5185241147774283e-05,
+      "loss": 0.6343,
+      "step": 4469
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0100118016035045,
+      "learning_rate": 1.5183092492025772e-05,
+      "loss": 0.5398,
+      "step": 4470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2192429254861237,
+      "learning_rate": 1.518094350904189e-05,
+      "loss": 0.5893,
+      "step": 4471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0759556270937056,
+      "learning_rate": 1.5178794198958313e-05,
+      "loss": 0.5805,
+      "step": 4472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2030424114466285,
+      "learning_rate": 1.517664456191074e-05,
+      "loss": 0.6144,
+      "step": 4473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2887942873192972,
+      "learning_rate": 1.5174494598034889e-05,
+      "loss": 0.6429,
+      "step": 4474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2304358583928532,
+      "learning_rate": 1.5172344307466493e-05,
+      "loss": 0.6387,
+      "step": 4475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.210997106254138,
+      "learning_rate": 1.517019369034132e-05,
+      "loss": 0.6773,
+      "step": 4476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1317197066917521,
+      "learning_rate": 1.516804274679514e-05,
+      "loss": 0.573,
+      "step": 4477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.226489799920522,
+      "learning_rate": 1.5165891476963763e-05,
+      "loss": 0.5784,
+      "step": 4478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.098668673459862,
+      "learning_rate": 1.5163739880983002e-05,
+      "loss": 0.5126,
+      "step": 4479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2200222222070949,
+      "learning_rate": 1.5161587958988699e-05,
+      "loss": 0.6562,
+      "step": 4480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1494085656928357,
+      "learning_rate": 1.515943571111672e-05,
+      "loss": 0.6214,
+      "step": 4481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0634176554973584,
+      "learning_rate": 1.5157283137502944e-05,
+      "loss": 0.5118,
+      "step": 4482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1172460260600112,
+      "learning_rate": 1.5155130238283277e-05,
+      "loss": 0.5477,
+      "step": 4483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.119030479350138,
+      "learning_rate": 1.5152977013593643e-05,
+      "loss": 0.5482,
+      "step": 4484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2970108972149006,
+      "learning_rate": 1.5150823463569979e-05,
+      "loss": 0.5596,
+      "step": 4485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1490856599276307,
+      "learning_rate": 1.514866958834826e-05,
+      "loss": 0.5687,
+      "step": 4486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1965998280984946,
+      "learning_rate": 1.5146515388064463e-05,
+      "loss": 0.5549,
+      "step": 4487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1337593593862998,
+      "learning_rate": 1.5144360862854597e-05,
+      "loss": 0.564,
+      "step": 4488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1684703961639018,
+      "learning_rate": 1.5142206012854693e-05,
+      "loss": 0.6251,
+      "step": 4489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1579033140327963,
+      "learning_rate": 1.5140050838200786e-05,
+      "loss": 0.6207,
+      "step": 4490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3167438470375483,
+      "learning_rate": 1.5137895339028955e-05,
+      "loss": 0.599,
+      "step": 4491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2358226737250553,
+      "learning_rate": 1.5135739515475281e-05,
+      "loss": 0.523,
+      "step": 4492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2380342447755142,
+      "learning_rate": 1.5133583367675878e-05,
+      "loss": 0.5884,
+      "step": 4493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2101236377161075,
+      "learning_rate": 1.5131426895766868e-05,
+      "loss": 0.5678,
+      "step": 4494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2288445303830322,
+      "learning_rate": 1.5129270099884403e-05,
+      "loss": 0.5841,
+      "step": 4495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.54702295693919,
+      "learning_rate": 1.5127112980164655e-05,
+      "loss": 0.6066,
+      "step": 4496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.113243236227367,
+      "learning_rate": 1.512495553674381e-05,
+      "loss": 0.591,
+      "step": 4497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1566739207341692,
+      "learning_rate": 1.5122797769758081e-05,
+      "loss": 0.5457,
+      "step": 4498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0650606154264042,
+      "learning_rate": 1.5120639679343702e-05,
+      "loss": 0.5679,
+      "step": 4499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1669668027279647,
+      "learning_rate": 1.5118481265636917e-05,
+      "loss": 0.5665,
+      "step": 4500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2619587106131802,
+      "learning_rate": 1.5116322528774005e-05,
+      "loss": 0.5949,
+      "step": 4501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2086465144706862,
+      "learning_rate": 1.5114163468891252e-05,
+      "loss": 0.5736,
+      "step": 4502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.182223795596568,
+      "learning_rate": 1.5112004086124976e-05,
+      "loss": 0.5736,
+      "step": 4503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1786402418021102,
+      "learning_rate": 1.5109844380611506e-05,
+      "loss": 0.5935,
+      "step": 4504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2747217267363526,
+      "learning_rate": 1.51076843524872e-05,
+      "loss": 0.5938,
+      "step": 4505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2519558863378772,
+      "learning_rate": 1.5105524001888425e-05,
+      "loss": 0.6118,
+      "step": 4506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2270405317698923,
+      "learning_rate": 1.510336332895158e-05,
+      "loss": 0.6242,
+      "step": 4507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1195539473517495,
+      "learning_rate": 1.5101202333813078e-05,
+      "loss": 0.5322,
+      "step": 4508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1557193775745187,
+      "learning_rate": 1.5099041016609355e-05,
+      "loss": 0.5888,
+      "step": 4509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1709091784165144,
+      "learning_rate": 1.5096879377476864e-05,
+      "loss": 0.5435,
+      "step": 4510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1386916615253955,
+      "learning_rate": 1.509471741655208e-05,
+      "loss": 0.5176,
+      "step": 4511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.200871326093473,
+      "learning_rate": 1.5092555133971502e-05,
+      "loss": 0.6216,
+      "step": 4512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2124918652291095,
+      "learning_rate": 1.5090392529871645e-05,
+      "loss": 0.5986,
+      "step": 4513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1511681200994655,
+      "learning_rate": 1.5088229604389045e-05,
+      "loss": 0.5848,
+      "step": 4514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2295573901278383,
+      "learning_rate": 1.5086066357660255e-05,
+      "loss": 0.5798,
+      "step": 4515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0988644504068925,
+      "learning_rate": 1.5083902789821854e-05,
+      "loss": 0.6043,
+      "step": 4516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.225616819440978,
+      "learning_rate": 1.5081738901010446e-05,
+      "loss": 0.5765,
+      "step": 4517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1822751698064238,
+      "learning_rate": 1.507957469136264e-05,
+      "loss": 0.6063,
+      "step": 4518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.113194244732716,
+      "learning_rate": 1.5077410161015078e-05,
+      "loss": 0.5454,
+      "step": 4519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2642688785506542,
+      "learning_rate": 1.5075245310104414e-05,
+      "loss": 0.6169,
+      "step": 4520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1296981621076856,
+      "learning_rate": 1.507308013876733e-05,
+      "loss": 0.5632,
+      "step": 4521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2368674405152718,
+      "learning_rate": 1.5070914647140522e-05,
+      "loss": 0.6307,
+      "step": 4522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2112597129125278,
+      "learning_rate": 1.5068748835360713e-05,
+      "loss": 0.5861,
+      "step": 4523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.219331749208178,
+      "learning_rate": 1.5066582703564638e-05,
+      "loss": 0.6005,
+      "step": 4524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2052131960459438,
+      "learning_rate": 1.5064416251889053e-05,
+      "loss": 0.6435,
+      "step": 4525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.123783566543255,
+      "learning_rate": 1.5062249480470742e-05,
+      "loss": 0.5214,
+      "step": 4526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1489162355505493,
+      "learning_rate": 1.5060082389446509e-05,
+      "loss": 0.5995,
+      "step": 4527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1174787128477695,
+      "learning_rate": 1.5057914978953166e-05,
+      "loss": 0.5736,
+      "step": 4528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1073052884852255,
+      "learning_rate": 1.5055747249127552e-05,
+      "loss": 0.611,
+      "step": 4529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0962024985518966,
+      "learning_rate": 1.5053579200106531e-05,
+      "loss": 0.5769,
+      "step": 4530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.175217756425074,
+      "learning_rate": 1.505141083202698e-05,
+      "loss": 0.5486,
+      "step": 4531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1204931797124718,
+      "learning_rate": 1.5049242145025806e-05,
+      "loss": 0.6081,
+      "step": 4532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1347182981303137,
+      "learning_rate": 1.5047073139239922e-05,
+      "loss": 0.6092,
+      "step": 4533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2278380352344656,
+      "learning_rate": 1.5044903814806273e-05,
+      "loss": 0.5842,
+      "step": 4534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2165225777590694,
+      "learning_rate": 1.5042734171861815e-05,
+      "loss": 0.5976,
+      "step": 4535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.194584843173679,
+      "learning_rate": 1.5040564210543532e-05,
+      "loss": 0.5858,
+      "step": 4536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0587298879323264,
+      "learning_rate": 1.5038393930988426e-05,
+      "loss": 0.5562,
+      "step": 4537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2387121756405723,
+      "learning_rate": 1.5036223333333517e-05,
+      "loss": 0.6154,
+      "step": 4538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1080072056987036,
+      "learning_rate": 1.5034052417715846e-05,
+      "loss": 0.6061,
+      "step": 4539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.07299707475603,
+      "learning_rate": 1.503188118427247e-05,
+      "loss": 0.4944,
+      "step": 4540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1760255173328118,
+      "learning_rate": 1.5029709633140476e-05,
+      "loss": 0.5106,
+      "step": 4541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1956653448178554,
+      "learning_rate": 1.5027537764456963e-05,
+      "loss": 0.5875,
+      "step": 4542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1648337862035085,
+      "learning_rate": 1.5025365578359053e-05,
+      "loss": 0.5466,
+      "step": 4543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.237878679848045,
+      "learning_rate": 1.5023193074983886e-05,
+      "loss": 0.6518,
+      "step": 4544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.07618068130684,
+      "learning_rate": 1.5021020254468623e-05,
+      "loss": 0.5112,
+      "step": 4545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1895439727114845,
+      "learning_rate": 1.5018847116950445e-05,
+      "loss": 0.6323,
+      "step": 4546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.193688529863557,
+      "learning_rate": 1.5016673662566558e-05,
+      "loss": 0.5816,
+      "step": 4547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2262444539387036,
+      "learning_rate": 1.501449989145418e-05,
+      "loss": 0.6077,
+      "step": 4548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1452765528097446,
+      "learning_rate": 1.501232580375055e-05,
+      "loss": 0.6007,
+      "step": 4549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.173722857485436,
+      "learning_rate": 1.5010151399592934e-05,
+      "loss": 0.5908,
+      "step": 4550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.285188204939091,
+      "learning_rate": 1.500797667911861e-05,
+      "loss": 0.6214,
+      "step": 4551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.233078384253019,
+      "learning_rate": 1.5005801642464879e-05,
+      "loss": 0.57,
+      "step": 4552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1913349568032714,
+      "learning_rate": 1.5003626289769066e-05,
+      "loss": 0.6218,
+      "step": 4553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1941912010496085,
+      "learning_rate": 1.5001450621168507e-05,
+      "loss": 0.5583,
+      "step": 4554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.332888106675841,
+      "learning_rate": 1.4999274636800572e-05,
+      "loss": 0.6261,
+      "step": 4555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1712951751595282,
+      "learning_rate": 1.4997098336802631e-05,
+      "loss": 0.5899,
+      "step": 4556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0782279919068292,
+      "learning_rate": 1.4994921721312092e-05,
+      "loss": 0.5431,
+      "step": 4557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.128441790961086,
+      "learning_rate": 1.4992744790466376e-05,
+      "loss": 0.5064,
+      "step": 4558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3008492025319331,
+      "learning_rate": 1.4990567544402918e-05,
+      "loss": 0.6312,
+      "step": 4559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.18534670904363,
+      "learning_rate": 1.4988389983259188e-05,
+      "loss": 0.537,
+      "step": 4560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1484538485862017,
+      "learning_rate": 1.4986212107172658e-05,
+      "loss": 0.5871,
+      "step": 4561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1217149349349158,
+      "learning_rate": 1.4984033916280833e-05,
+      "loss": 0.5938,
+      "step": 4562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1819202943350071,
+      "learning_rate": 1.4981855410721236e-05,
+      "loss": 0.5871,
+      "step": 4563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1638147999457185,
+      "learning_rate": 1.4979676590631398e-05,
+      "loss": 0.6018,
+      "step": 4564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1778014813942903,
+      "learning_rate": 1.4977497456148891e-05,
+      "loss": 0.5923,
+      "step": 4565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.099815986153892,
+      "learning_rate": 1.4975318007411284e-05,
+      "loss": 0.507,
+      "step": 4566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0380142775713128,
+      "learning_rate": 1.4973138244556184e-05,
+      "loss": 0.482,
+      "step": 4567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.217544203705804,
+      "learning_rate": 1.497095816772121e-05,
+      "loss": 0.6523,
+      "step": 4568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1779864849552317,
+      "learning_rate": 1.4968777777043997e-05,
+      "loss": 0.5472,
+      "step": 4569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.189498424450415,
+      "learning_rate": 1.496659707266221e-05,
+      "loss": 0.538,
+      "step": 4570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.19365597299684,
+      "learning_rate": 1.4964416054713525e-05,
+      "loss": 0.5788,
+      "step": 4571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2080286836787093,
+      "learning_rate": 1.4962234723335642e-05,
+      "loss": 0.5859,
+      "step": 4572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1123784738236282,
+      "learning_rate": 1.4960053078666278e-05,
+      "loss": 0.5111,
+      "step": 4573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.134893886126242,
+      "learning_rate": 1.4957871120843172e-05,
+      "loss": 0.5944,
+      "step": 4574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1983506729523272,
+      "learning_rate": 1.4955688850004087e-05,
+      "loss": 0.5461,
+      "step": 4575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2331431072469832,
+      "learning_rate": 1.495350626628679e-05,
+      "loss": 0.6081,
+      "step": 4576
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.217480756619316,
+      "learning_rate": 1.4951323369829091e-05,
+      "loss": 0.5024,
+      "step": 4577
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.130900275194096,
+      "learning_rate": 1.4949140160768803e-05,
+      "loss": 0.5241,
+      "step": 4578
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0714250893763317,
+      "learning_rate": 1.4946956639243757e-05,
+      "loss": 0.5332,
+      "step": 4579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1859485128140161,
+      "learning_rate": 1.4944772805391821e-05,
+      "loss": 0.5833,
+      "step": 4580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1704996812405288,
+      "learning_rate": 1.4942588659350863e-05,
+      "loss": 0.5798,
+      "step": 4581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1671876879261687,
+      "learning_rate": 1.4940404201258782e-05,
+      "loss": 0.5772,
+      "step": 4582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2348541102553379,
+      "learning_rate": 1.4938219431253499e-05,
+      "loss": 0.6385,
+      "step": 4583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1598463061886657,
+      "learning_rate": 1.4936034349472941e-05,
+      "loss": 0.5616,
+      "step": 4584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1316569781743098,
+      "learning_rate": 1.4933848956055068e-05,
+      "loss": 0.5603,
+      "step": 4585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1578602789983201,
+      "learning_rate": 1.4931663251137856e-05,
+      "loss": 0.5509,
+      "step": 4586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0901857885334052,
+      "learning_rate": 1.4929477234859299e-05,
+      "loss": 0.5413,
+      "step": 4587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1502188536504823,
+      "learning_rate": 1.4927290907357415e-05,
+      "loss": 0.5587,
+      "step": 4588
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2573429439594714,
+      "learning_rate": 1.4925104268770227e-05,
+      "loss": 0.6477,
+      "step": 4589
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0957070416875185,
+      "learning_rate": 1.4922917319235804e-05,
+      "loss": 0.5368,
+      "step": 4590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1902997491283291,
+      "learning_rate": 1.4920730058892205e-05,
+      "loss": 0.585,
+      "step": 4591
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2319059183623426,
+      "learning_rate": 1.4918542487877535e-05,
+      "loss": 0.592,
+      "step": 4592
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1735340327596926,
+      "learning_rate": 1.49163546063299e-05,
+      "loss": 0.5087,
+      "step": 4593
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2132498045000901,
+      "learning_rate": 1.4914166414387433e-05,
+      "loss": 0.5851,
+      "step": 4594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2265314960813345,
+      "learning_rate": 1.4911977912188284e-05,
+      "loss": 0.5568,
+      "step": 4595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0343460078944768,
+      "learning_rate": 1.490978909987063e-05,
+      "loss": 0.5331,
+      "step": 4596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2338777820307305,
+      "learning_rate": 1.4907599977572659e-05,
+      "loss": 0.586,
+      "step": 4597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1826844200219562,
+      "learning_rate": 1.490541054543258e-05,
+      "loss": 0.5681,
+      "step": 4598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2097961954030196,
+      "learning_rate": 1.4903220803588627e-05,
+      "loss": 0.5671,
+      "step": 4599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2305487228320586,
+      "learning_rate": 1.4901030752179044e-05,
+      "loss": 0.6165,
+      "step": 4600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2404697945702465,
+      "learning_rate": 1.4898840391342107e-05,
+      "loss": 0.5561,
+      "step": 4601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1278644757596286,
+      "learning_rate": 1.4896649721216101e-05,
+      "loss": 0.5668,
+      "step": 4602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0960607370093876,
+      "learning_rate": 1.4894458741939333e-05,
+      "loss": 0.585,
+      "step": 4603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1829674199794296,
+      "learning_rate": 1.4892267453650133e-05,
+      "loss": 0.5935,
+      "step": 4604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0172343487945301,
+      "learning_rate": 1.4890075856486848e-05,
+      "loss": 0.4999,
+      "step": 4605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2785751986238711,
+      "learning_rate": 1.4887883950587845e-05,
+      "loss": 0.5755,
+      "step": 4606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2564671115297674,
+      "learning_rate": 1.488569173609151e-05,
+      "loss": 0.5793,
+      "step": 4607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2780503118279807,
+      "learning_rate": 1.488349921313625e-05,
+      "loss": 0.5999,
+      "step": 4608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1265165914845163,
+      "learning_rate": 1.4881306381860485e-05,
+      "loss": 0.5823,
+      "step": 4609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1812260176227336,
+      "learning_rate": 1.4879113242402668e-05,
+      "loss": 0.5505,
+      "step": 4610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1186908663673416,
+      "learning_rate": 1.4876919794901256e-05,
+      "loss": 0.5666,
+      "step": 4611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1935460620456722,
+      "learning_rate": 1.487472603949474e-05,
+      "loss": 0.5874,
+      "step": 4612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2067492998309621,
+      "learning_rate": 1.4872531976321619e-05,
+      "loss": 0.655,
+      "step": 4613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.104979174145839,
+      "learning_rate": 1.4870337605520408e-05,
+      "loss": 0.5547,
+      "step": 4614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0634184962482276,
+      "learning_rate": 1.4868142927229662e-05,
+      "loss": 0.5519,
+      "step": 4615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1808072264447613,
+      "learning_rate": 1.4865947941587938e-05,
+      "loss": 0.6282,
+      "step": 4616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2570182235355767,
+      "learning_rate": 1.4863752648733812e-05,
+      "loss": 0.5545,
+      "step": 4617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2856973361230202,
+      "learning_rate": 1.486155704880589e-05,
+      "loss": 0.675,
+      "step": 4618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3064167646339222,
+      "learning_rate": 1.4859361141942788e-05,
+      "loss": 0.6257,
+      "step": 4619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1636661136275848,
+      "learning_rate": 1.4857164928283143e-05,
+      "loss": 0.5525,
+      "step": 4620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2044451023162963,
+      "learning_rate": 1.4854968407965621e-05,
+      "loss": 0.589,
+      "step": 4621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4738054319768479,
+      "learning_rate": 1.4852771581128895e-05,
+      "loss": 0.6303,
+      "step": 4622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1010681558175648,
+      "learning_rate": 1.4850574447911661e-05,
+      "loss": 0.5883,
+      "step": 4623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1235184450659916,
+      "learning_rate": 1.4848377008452635e-05,
+      "loss": 0.5364,
+      "step": 4624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2718744338875525,
+      "learning_rate": 1.4846179262890554e-05,
+      "loss": 0.6048,
+      "step": 4625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.236600294482748,
+      "learning_rate": 1.4843981211364175e-05,
+      "loss": 0.6051,
+      "step": 4626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1642824515615444,
+      "learning_rate": 1.484178285401227e-05,
+      "loss": 0.5463,
+      "step": 4627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2228783353986021,
+      "learning_rate": 1.4839584190973633e-05,
+      "loss": 0.5762,
+      "step": 4628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1734859837498324,
+      "learning_rate": 1.4837385222387078e-05,
+      "loss": 0.5542,
+      "step": 4629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0824503723820524,
+      "learning_rate": 1.4835185948391433e-05,
+      "loss": 0.6399,
+      "step": 4630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1092326650513769,
+      "learning_rate": 1.4832986369125558e-05,
+      "loss": 0.5604,
+      "step": 4631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3168809527247678,
+      "learning_rate": 1.4830786484728315e-05,
+      "loss": 0.6606,
+      "step": 4632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1675323385460907,
+      "learning_rate": 1.4828586295338597e-05,
+      "loss": 0.5924,
+      "step": 4633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1717783061507685,
+      "learning_rate": 1.4826385801095315e-05,
+      "loss": 0.5686,
+      "step": 4634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.197491174956039,
+      "learning_rate": 1.4824185002137396e-05,
+      "loss": 0.5732,
+      "step": 4635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.130614997140509,
+      "learning_rate": 1.4821983898603791e-05,
+      "loss": 0.5606,
+      "step": 4636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1701476527237085,
+      "learning_rate": 1.4819782490633463e-05,
+      "loss": 0.5743,
+      "step": 4637
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1435780380301737,
+      "learning_rate": 1.4817580778365396e-05,
+      "loss": 0.5634,
+      "step": 4638
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2038043692601252,
+      "learning_rate": 1.4815378761938603e-05,
+      "loss": 0.6005,
+      "step": 4639
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1275125632252232,
+      "learning_rate": 1.4813176441492104e-05,
+      "loss": 0.5623,
+      "step": 4640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1594266813126217,
+      "learning_rate": 1.4810973817164941e-05,
+      "loss": 0.5531,
+      "step": 4641
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2656768505636797,
+      "learning_rate": 1.4808770889096184e-05,
+      "loss": 0.6042,
+      "step": 4642
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1705646564105916,
+      "learning_rate": 1.4806567657424908e-05,
+      "loss": 0.5571,
+      "step": 4643
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1688436338075159,
+      "learning_rate": 1.4804364122290217e-05,
+      "loss": 0.5717,
+      "step": 4644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.200153346992501,
+      "learning_rate": 1.4802160283831233e-05,
+      "loss": 0.5803,
+      "step": 4645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2668605942511664,
+      "learning_rate": 1.4799956142187094e-05,
+      "loss": 0.5679,
+      "step": 4646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3013203279983399,
+      "learning_rate": 1.479775169749696e-05,
+      "loss": 0.6632,
+      "step": 4647
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.156828297517851,
+      "learning_rate": 1.4795546949900006e-05,
+      "loss": 0.5417,
+      "step": 4648
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1721538466257775,
+      "learning_rate": 1.4793341899535434e-05,
+      "loss": 0.5572,
+      "step": 4649
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2127220540378292,
+      "learning_rate": 1.4791136546542454e-05,
+      "loss": 0.5855,
+      "step": 4650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2433614881223698,
+      "learning_rate": 1.4788930891060307e-05,
+      "loss": 0.6232,
+      "step": 4651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1928735921164484,
+      "learning_rate": 1.4786724933228247e-05,
+      "loss": 0.6087,
+      "step": 4652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.23211027580133,
+      "learning_rate": 1.4784518673185542e-05,
+      "loss": 0.6004,
+      "step": 4653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.157972907976523,
+      "learning_rate": 1.478231211107149e-05,
+      "loss": 0.5712,
+      "step": 4654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1700557574818624,
+      "learning_rate": 1.47801052470254e-05,
+      "loss": 0.6394,
+      "step": 4655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2900370295333148,
+      "learning_rate": 1.4777898081186606e-05,
+      "loss": 0.6146,
+      "step": 4656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2731685705505695,
+      "learning_rate": 1.4775690613694453e-05,
+      "loss": 0.6077,
+      "step": 4657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.286985631618011,
+      "learning_rate": 1.4773482844688313e-05,
+      "loss": 0.6205,
+      "step": 4658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0994185016086377,
+      "learning_rate": 1.4771274774307573e-05,
+      "loss": 0.5765,
+      "step": 4659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2076429280106338,
+      "learning_rate": 1.4769066402691641e-05,
+      "loss": 0.6024,
+      "step": 4660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2295042102519795,
+      "learning_rate": 1.476685772997994e-05,
+      "loss": 0.6074,
+      "step": 4661
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1001036161692197,
+      "learning_rate": 1.476464875631192e-05,
+      "loss": 0.5811,
+      "step": 4662
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3983361830850933,
+      "learning_rate": 1.4762439481827038e-05,
+      "loss": 0.6713,
+      "step": 4663
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1104922511634667,
+      "learning_rate": 1.4760229906664782e-05,
+      "loss": 0.5738,
+      "step": 4664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1672524389460859,
+      "learning_rate": 1.4758020030964653e-05,
+      "loss": 0.6111,
+      "step": 4665
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2111903265080481,
+      "learning_rate": 1.4755809854866172e-05,
+      "loss": 0.6154,
+      "step": 4666
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0708957948292044,
+      "learning_rate": 1.4753599378508876e-05,
+      "loss": 0.5897,
+      "step": 4667
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1909672162711045,
+      "learning_rate": 1.4751388602032326e-05,
+      "loss": 0.585,
+      "step": 4668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3186552492787873,
+      "learning_rate": 1.4749177525576102e-05,
+      "loss": 0.5289,
+      "step": 4669
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2254133247313488,
+      "learning_rate": 1.4746966149279796e-05,
+      "loss": 0.5357,
+      "step": 4670
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.212856765249875,
+      "learning_rate": 1.4744754473283024e-05,
+      "loss": 0.6066,
+      "step": 4671
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2209873204334436,
+      "learning_rate": 1.4742542497725428e-05,
+      "loss": 0.5713,
+      "step": 4672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1066103052145848,
+      "learning_rate": 1.4740330222746653e-05,
+      "loss": 0.5507,
+      "step": 4673
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.056525938117344,
+      "learning_rate": 1.4738117648486375e-05,
+      "loss": 0.5389,
+      "step": 4674
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.242217777291066,
+      "learning_rate": 1.473590477508428e-05,
+      "loss": 0.6362,
+      "step": 4675
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1351115607163516,
+      "learning_rate": 1.4733691602680088e-05,
+      "loss": 0.61,
+      "step": 4676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.10468385687717,
+      "learning_rate": 1.4731478131413519e-05,
+      "loss": 0.5112,
+      "step": 4677
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2354500350605704,
+      "learning_rate": 1.4729264361424325e-05,
+      "loss": 0.5876,
+      "step": 4678
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1107857022348495,
+      "learning_rate": 1.4727050292852272e-05,
+      "loss": 0.6043,
+      "step": 4679
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1257512445309525,
+      "learning_rate": 1.4724835925837146e-05,
+      "loss": 0.5526,
+      "step": 4680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3035727667008492,
+      "learning_rate": 1.4722621260518752e-05,
+      "loss": 0.5747,
+      "step": 4681
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0787942440114195,
+      "learning_rate": 1.4720406297036913e-05,
+      "loss": 0.5057,
+      "step": 4682
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2150950279279153,
+      "learning_rate": 1.4718191035531468e-05,
+      "loss": 0.5375,
+      "step": 4683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2089244229884872,
+      "learning_rate": 1.471597547614228e-05,
+      "loss": 0.5637,
+      "step": 4684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.140700925625432,
+      "learning_rate": 1.471375961900923e-05,
+      "loss": 0.5832,
+      "step": 4685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1614544216939748,
+      "learning_rate": 1.4711543464272218e-05,
+      "loss": 0.5573,
+      "step": 4686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2904417577530933,
+      "learning_rate": 1.4709327012071157e-05,
+      "loss": 0.5754,
+      "step": 4687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1884137202052194,
+      "learning_rate": 1.4707110262545983e-05,
+      "loss": 0.5735,
+      "step": 4688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3137574303268986,
+      "learning_rate": 1.4704893215836653e-05,
+      "loss": 0.6075,
+      "step": 4689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1424801515486995,
+      "learning_rate": 1.4702675872083141e-05,
+      "loss": 0.591,
+      "step": 4690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0066994723496627,
+      "learning_rate": 1.470045823142544e-05,
+      "loss": 0.5258,
+      "step": 4691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0826355386328324,
+      "learning_rate": 1.469824029400356e-05,
+      "loss": 0.554,
+      "step": 4692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.171799059632538,
+      "learning_rate": 1.4696022059957527e-05,
+      "loss": 0.6025,
+      "step": 4693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1649420059078481,
+      "learning_rate": 1.4693803529427393e-05,
+      "loss": 0.5831,
+      "step": 4694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2874050234975287,
+      "learning_rate": 1.469158470255323e-05,
+      "loss": 0.5842,
+      "step": 4695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1068601445084842,
+      "learning_rate": 1.4689365579475117e-05,
+      "loss": 0.5143,
+      "step": 4696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2106024001596387,
+      "learning_rate": 1.4687146160333162e-05,
+      "loss": 0.582,
+      "step": 4697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1885202943628286,
+      "learning_rate": 1.4684926445267485e-05,
+      "loss": 0.5579,
+      "step": 4698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.265046258163542,
+      "learning_rate": 1.4682706434418229e-05,
+      "loss": 0.6054,
+      "step": 4699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1653156064217927,
+      "learning_rate": 1.468048612792556e-05,
+      "loss": 0.6027,
+      "step": 4700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2136577446995616,
+      "learning_rate": 1.467826552592965e-05,
+      "loss": 0.584,
+      "step": 4701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1824699581112545,
+      "learning_rate": 1.4676044628570707e-05,
+      "loss": 0.5522,
+      "step": 4702
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9981257338452493,
+      "learning_rate": 1.4673823435988933e-05,
+      "loss": 0.4888,
+      "step": 4703
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2072119716422292,
+      "learning_rate": 1.4671601948324577e-05,
+      "loss": 0.5952,
+      "step": 4704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1882638983929963,
+      "learning_rate": 1.4669380165717889e-05,
+      "loss": 0.6087,
+      "step": 4705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.216398121752131,
+      "learning_rate": 1.4667158088309137e-05,
+      "loss": 0.6106,
+      "step": 4706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.187912919381898,
+      "learning_rate": 1.4664935716238615e-05,
+      "loss": 0.5989,
+      "step": 4707
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1497304476083048,
+      "learning_rate": 1.4662713049646637e-05,
+      "loss": 0.6008,
+      "step": 4708
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.152337905093095,
+      "learning_rate": 1.4660490088673525e-05,
+      "loss": 0.56,
+      "step": 4709
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1848559306485862,
+      "learning_rate": 1.4658266833459629e-05,
+      "loss": 0.5928,
+      "step": 4710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2017497835676911,
+      "learning_rate": 1.4656043284145316e-05,
+      "loss": 0.5771,
+      "step": 4711
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1954164646787175,
+      "learning_rate": 1.4653819440870965e-05,
+      "loss": 0.5681,
+      "step": 4712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0869511326366013,
+      "learning_rate": 1.4651595303776986e-05,
+      "loss": 0.5519,
+      "step": 4713
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0623309618054928,
+      "learning_rate": 1.4649370873003794e-05,
+      "loss": 0.5548,
+      "step": 4714
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2850538404295495,
+      "learning_rate": 1.4647146148691831e-05,
+      "loss": 0.6135,
+      "step": 4715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1221894017593557,
+      "learning_rate": 1.4644921130981558e-05,
+      "loss": 0.5488,
+      "step": 4716
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1878619395351635,
+      "learning_rate": 1.4642695820013446e-05,
+      "loss": 0.5899,
+      "step": 4717
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2727229448034345,
+      "learning_rate": 1.4640470215927998e-05,
+      "loss": 0.6163,
+      "step": 4718
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.28863749578199,
+      "learning_rate": 1.463824431886572e-05,
+      "loss": 0.602,
+      "step": 4719
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.12852978616599,
+      "learning_rate": 1.4636018128967149e-05,
+      "loss": 0.5947,
+      "step": 4720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2562087835126583,
+      "learning_rate": 1.4633791646372837e-05,
+      "loss": 0.6434,
+      "step": 4721
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.373995761009685,
+      "learning_rate": 1.4631564871223346e-05,
+      "loss": 0.5608,
+      "step": 4722
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.223947982203224,
+      "learning_rate": 1.4629337803659274e-05,
+      "loss": 0.5634,
+      "step": 4723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1323658029570405,
+      "learning_rate": 1.4627110443821217e-05,
+      "loss": 0.6322,
+      "step": 4724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1576216011355829,
+      "learning_rate": 1.462488279184981e-05,
+      "loss": 0.5958,
+      "step": 4725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1545356722938178,
+      "learning_rate": 1.4622654847885688e-05,
+      "loss": 0.5311,
+      "step": 4726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1210583405435337,
+      "learning_rate": 1.4620426612069519e-05,
+      "loss": 0.5077,
+      "step": 4727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1855321942949153,
+      "learning_rate": 1.4618198084541977e-05,
+      "loss": 0.6277,
+      "step": 4728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.137618190787337,
+      "learning_rate": 1.4615969265443762e-05,
+      "loss": 0.5715,
+      "step": 4729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.088656308405371,
+      "learning_rate": 1.4613740154915594e-05,
+      "loss": 0.497,
+      "step": 4730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1511257652986857,
+      "learning_rate": 1.4611510753098208e-05,
+      "loss": 0.6052,
+      "step": 4731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2175385738896742,
+      "learning_rate": 1.4609281060132352e-05,
+      "loss": 0.5224,
+      "step": 4732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3204040495603833,
+      "learning_rate": 1.4607051076158805e-05,
+      "loss": 0.6724,
+      "step": 4733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2755219886974971,
+      "learning_rate": 1.4604820801318351e-05,
+      "loss": 0.5868,
+      "step": 4734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0580284608174968,
+      "learning_rate": 1.4602590235751806e-05,
+      "loss": 0.5534,
+      "step": 4735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1920072946805294,
+      "learning_rate": 1.4600359379599992e-05,
+      "loss": 0.5588,
+      "step": 4736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2067158110939797,
+      "learning_rate": 1.4598128233003754e-05,
+      "loss": 0.6165,
+      "step": 4737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1458799988462935,
+      "learning_rate": 1.4595896796103959e-05,
+      "loss": 0.5861,
+      "step": 4738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2734813565618301,
+      "learning_rate": 1.4593665069041484e-05,
+      "loss": 0.5554,
+      "step": 4739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0933440953884168,
+      "learning_rate": 1.4591433051957237e-05,
+      "loss": 0.5109,
+      "step": 4740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.079024064401115,
+      "learning_rate": 1.4589200744992134e-05,
+      "loss": 0.5394,
+      "step": 4741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1387230156591666,
+      "learning_rate": 1.4586968148287106e-05,
+      "loss": 0.5086,
+      "step": 4742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1098926035158136,
+      "learning_rate": 1.4584735261983118e-05,
+      "loss": 0.5804,
+      "step": 4743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2295573901278383,
+      "learning_rate": 1.4582502086221136e-05,
+      "loss": 0.5668,
+      "step": 4744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1761754791295764,
+      "learning_rate": 1.4580268621142155e-05,
+      "loss": 0.5156,
+      "step": 4745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1185849925845492,
+      "learning_rate": 1.4578034866887186e-05,
+      "loss": 0.5891,
+      "step": 4746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.169121979865239,
+      "learning_rate": 1.4575800823597255e-05,
+      "loss": 0.5588,
+      "step": 4747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2061909339951176,
+      "learning_rate": 1.4573566491413409e-05,
+      "loss": 0.6009,
+      "step": 4748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2298542244132797,
+      "learning_rate": 1.4571331870476716e-05,
+      "loss": 0.6244,
+      "step": 4749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2345690514344285,
+      "learning_rate": 1.4569096960928255e-05,
+      "loss": 0.5856,
+      "step": 4750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.28085516451661,
+      "learning_rate": 1.4566861762909133e-05,
+      "loss": 0.596,
+      "step": 4751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1875397023038996,
+      "learning_rate": 1.456462627656046e-05,
+      "loss": 0.5694,
+      "step": 4752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2369710446810522,
+      "learning_rate": 1.4562390502023384e-05,
+      "loss": 0.6166,
+      "step": 4753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2610003903575175,
+      "learning_rate": 1.4560154439439056e-05,
+      "loss": 0.6153,
+      "step": 4754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1656802923674987,
+      "learning_rate": 1.4557918088948652e-05,
+      "loss": 0.5819,
+      "step": 4755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0626824166276674,
+      "learning_rate": 1.4555681450693365e-05,
+      "loss": 0.5174,
+      "step": 4756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0902383288288213,
+      "learning_rate": 1.45534445248144e-05,
+      "loss": 0.5509,
+      "step": 4757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2193939757324839,
+      "learning_rate": 1.4551207311452991e-05,
+      "loss": 0.5316,
+      "step": 4758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.202331281154888,
+      "learning_rate": 1.4548969810750382e-05,
+      "loss": 0.6397,
+      "step": 4759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1999417588882453,
+      "learning_rate": 1.454673202284784e-05,
+      "loss": 0.5375,
+      "step": 4760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1422843881574958,
+      "learning_rate": 1.4544493947886648e-05,
+      "loss": 0.5942,
+      "step": 4761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.162060956186703,
+      "learning_rate": 1.4542255586008105e-05,
+      "loss": 0.5818,
+      "step": 4762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1087109902281118,
+      "learning_rate": 1.4540016937353531e-05,
+      "loss": 0.5732,
+      "step": 4763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1766382686945234,
+      "learning_rate": 1.4537778002064268e-05,
+      "loss": 0.5504,
+      "step": 4764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1572959396233062,
+      "learning_rate": 1.4535538780281666e-05,
+      "loss": 0.5931,
+      "step": 4765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.191828718482068,
+      "learning_rate": 1.4533299272147103e-05,
+      "loss": 0.5953,
+      "step": 4766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1207880008610662,
+      "learning_rate": 1.4531059477801965e-05,
+      "loss": 0.5145,
+      "step": 4767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1240205739877187,
+      "learning_rate": 1.4528819397387663e-05,
+      "loss": 0.4981,
+      "step": 4768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1965528547645943,
+      "learning_rate": 1.4526579031045631e-05,
+      "loss": 0.6224,
+      "step": 4769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2179431689930071,
+      "learning_rate": 1.452433837891731e-05,
+      "loss": 0.6308,
+      "step": 4770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1330384127846598,
+      "learning_rate": 1.4522097441144166e-05,
+      "loss": 0.5509,
+      "step": 4771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1060478931011515,
+      "learning_rate": 1.4519856217867676e-05,
+      "loss": 0.5299,
+      "step": 4772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1436129586922932,
+      "learning_rate": 1.4517614709229345e-05,
+      "loss": 0.5858,
+      "step": 4773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.175322433628499,
+      "learning_rate": 1.451537291537069e-05,
+      "loss": 0.5401,
+      "step": 4774
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2540039784978272,
+      "learning_rate": 1.4513130836433247e-05,
+      "loss": 0.6495,
+      "step": 4775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2930560125770585,
+      "learning_rate": 1.451088847255857e-05,
+      "loss": 0.6149,
+      "step": 4776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.211144066696648,
+      "learning_rate": 1.4508645823888228e-05,
+      "loss": 0.6141,
+      "step": 4777
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1302796559601922,
+      "learning_rate": 1.4506402890563813e-05,
+      "loss": 0.5816,
+      "step": 4778
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2926972902345264,
+      "learning_rate": 1.4504159672726937e-05,
+      "loss": 0.6145,
+      "step": 4779
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.201346978551093,
+      "learning_rate": 1.4501916170519221e-05,
+      "loss": 0.5835,
+      "step": 4780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2820652949655547,
+      "learning_rate": 1.4499672384082312e-05,
+      "loss": 0.616,
+      "step": 4781
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.250056027110474,
+      "learning_rate": 1.4497428313557866e-05,
+      "loss": 0.6359,
+      "step": 4782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2056339904703068,
+      "learning_rate": 1.449518395908757e-05,
+      "loss": 0.581,
+      "step": 4783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0803945122823422,
+      "learning_rate": 1.4492939320813117e-05,
+      "loss": 0.5485,
+      "step": 4784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.097617749390821,
+      "learning_rate": 1.4490694398876228e-05,
+      "loss": 0.5644,
+      "step": 4785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2280019586870559,
+      "learning_rate": 1.448844919341863e-05,
+      "loss": 0.5822,
+      "step": 4786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1724804649271054,
+      "learning_rate": 1.4486203704582075e-05,
+      "loss": 0.5991,
+      "step": 4787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2577953455922122,
+      "learning_rate": 1.4483957932508338e-05,
+      "loss": 0.6295,
+      "step": 4788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3325087958183344,
+      "learning_rate": 1.4481711877339202e-05,
+      "loss": 0.66,
+      "step": 4789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1686819699321465,
+      "learning_rate": 1.447946553921647e-05,
+      "loss": 0.5606,
+      "step": 4790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1877135034669553,
+      "learning_rate": 1.4477218918281967e-05,
+      "loss": 0.5913,
+      "step": 4791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0846278574643111,
+      "learning_rate": 1.4474972014677537e-05,
+      "loss": 0.5443,
+      "step": 4792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1014566573569242,
+      "learning_rate": 1.4472724828545035e-05,
+      "loss": 0.5775,
+      "step": 4793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0747883344345852,
+      "learning_rate": 1.447047736002634e-05,
+      "loss": 0.5655,
+      "step": 4794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2638569951254384,
+      "learning_rate": 1.4468229609263343e-05,
+      "loss": 0.592,
+      "step": 4795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0124472569044305,
+      "learning_rate": 1.4465981576397957e-05,
+      "loss": 0.5514,
+      "step": 4796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3247691223238744,
+      "learning_rate": 1.4463733261572114e-05,
+      "loss": 0.6501,
+      "step": 4797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1495173042491573,
+      "learning_rate": 1.4461484664927758e-05,
+      "loss": 0.6023,
+      "step": 4798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1215567347223456,
+      "learning_rate": 1.4459235786606861e-05,
+      "loss": 0.5399,
+      "step": 4799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2231811274309439,
+      "learning_rate": 1.44569866267514e-05,
+      "loss": 0.5888,
+      "step": 4800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2151362812078543,
+      "learning_rate": 1.4454737185503375e-05,
+      "loss": 0.5955,
+      "step": 4801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.082043864593238,
+      "learning_rate": 1.4452487463004815e-05,
+      "loss": 0.5976,
+      "step": 4802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0975717532396327,
+      "learning_rate": 1.4450237459397742e-05,
+      "loss": 0.5471,
+      "step": 4803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2611125517280033,
+      "learning_rate": 1.4447987174824225e-05,
+      "loss": 0.5977,
+      "step": 4804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.042119970235621,
+      "learning_rate": 1.4445736609426324e-05,
+      "loss": 0.5044,
+      "step": 4805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2066174634277038,
+      "learning_rate": 1.4443485763346135e-05,
+      "loss": 0.6184,
+      "step": 4806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.081636763180403,
+      "learning_rate": 1.4441234636725767e-05,
+      "loss": 0.5276,
+      "step": 4807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1733889654817238,
+      "learning_rate": 1.4438983229707338e-05,
+      "loss": 0.5471,
+      "step": 4808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.165812258980968,
+      "learning_rate": 1.4436731542433e-05,
+      "loss": 0.5866,
+      "step": 4809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.277588847032594,
+      "learning_rate": 1.4434479575044908e-05,
+      "loss": 0.5726,
+      "step": 4810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3050374001591714,
+      "learning_rate": 1.443222732768524e-05,
+      "loss": 0.6454,
+      "step": 4811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.185846331764837,
+      "learning_rate": 1.4429974800496194e-05,
+      "loss": 0.563,
+      "step": 4812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.205603239378219,
+      "learning_rate": 1.4427721993619983e-05,
+      "loss": 0.5707,
+      "step": 4813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2266679949252222,
+      "learning_rate": 1.4425468907198843e-05,
+      "loss": 0.5924,
+      "step": 4814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.149111387837983,
+      "learning_rate": 1.4423215541375013e-05,
+      "loss": 0.5273,
+      "step": 4815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1426260652836573,
+      "learning_rate": 1.4420961896290764e-05,
+      "loss": 0.5476,
+      "step": 4816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.303687711930291,
+      "learning_rate": 1.4418707972088386e-05,
+      "loss": 0.7074,
+      "step": 4817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.138059840926854,
+      "learning_rate": 1.4416453768910173e-05,
+      "loss": 0.5787,
+      "step": 4818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1330971721886915,
+      "learning_rate": 1.4414199286898449e-05,
+      "loss": 0.5522,
+      "step": 4819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2630363646925478,
+      "learning_rate": 1.4411944526195551e-05,
+      "loss": 0.608,
+      "step": 4820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1907944906550014,
+      "learning_rate": 1.4409689486943829e-05,
+      "loss": 0.5857,
+      "step": 4821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2335958315328632,
+      "learning_rate": 1.4407434169285664e-05,
+      "loss": 0.5955,
+      "step": 4822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1670974492439778,
+      "learning_rate": 1.4405178573363435e-05,
+      "loss": 0.5905,
+      "step": 4823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.194352506209174,
+      "learning_rate": 1.4402922699319557e-05,
+      "loss": 0.5869,
+      "step": 4824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.102264180816152,
+      "learning_rate": 1.4400666547296456e-05,
+      "loss": 0.5743,
+      "step": 4825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0537866206384232,
+      "learning_rate": 1.4398410117436566e-05,
+      "loss": 0.4935,
+      "step": 4826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1581988535604466,
+      "learning_rate": 1.4396153409882356e-05,
+      "loss": 0.5872,
+      "step": 4827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1849326941936331,
+      "learning_rate": 1.4393896424776296e-05,
+      "loss": 0.6047,
+      "step": 4828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1805735917053748,
+      "learning_rate": 1.439163916226089e-05,
+      "loss": 0.5915,
+      "step": 4829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1801905790891936,
+      "learning_rate": 1.4389381622478644e-05,
+      "loss": 0.5934,
+      "step": 4830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1340017456027796,
+      "learning_rate": 1.438712380557209e-05,
+      "loss": 0.5643,
+      "step": 4831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1279649869341102,
+      "learning_rate": 1.4384865711683778e-05,
+      "loss": 0.5166,
+      "step": 4832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.099863189078922,
+      "learning_rate": 1.4382607340956265e-05,
+      "loss": 0.5906,
+      "step": 4833
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1926667596270755,
+      "learning_rate": 1.4380348693532144e-05,
+      "loss": 0.6094,
+      "step": 4834
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2270666167713262,
+      "learning_rate": 1.4378089769554009e-05,
+      "loss": 0.6197,
+      "step": 4835
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2928307220450972,
+      "learning_rate": 1.4375830569164478e-05,
+      "loss": 0.6853,
+      "step": 4836
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1981011566511857,
+      "learning_rate": 1.4373571092506189e-05,
+      "loss": 0.5279,
+      "step": 4837
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0756776093512608,
+      "learning_rate": 1.437131133972179e-05,
+      "loss": 0.4926,
+      "step": 4838
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.200482029263595,
+      "learning_rate": 1.4369051310953954e-05,
+      "loss": 0.5534,
+      "step": 4839
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1552876256360114,
+      "learning_rate": 1.436679100634537e-05,
+      "loss": 0.5862,
+      "step": 4840
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2420653277122962,
+      "learning_rate": 1.4364530426038734e-05,
+      "loss": 0.6291,
+      "step": 4841
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.114312908605802,
+      "learning_rate": 1.436226957017678e-05,
+      "loss": 0.5336,
+      "step": 4842
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1131179422093747,
+      "learning_rate": 1.436000843890224e-05,
+      "loss": 0.5634,
+      "step": 4843
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.042571545805598,
+      "learning_rate": 1.435774703235787e-05,
+      "loss": 0.5255,
+      "step": 4844
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1318485762883639,
+      "learning_rate": 1.4355485350686449e-05,
+      "loss": 0.5397,
+      "step": 4845
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2434428847034074,
+      "learning_rate": 1.4353223394030767e-05,
+      "loss": 0.6402,
+      "step": 4846
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1766315819874225,
+      "learning_rate": 1.4350961162533627e-05,
+      "loss": 0.5766,
+      "step": 4847
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2268830384716451,
+      "learning_rate": 1.434869865633787e-05,
+      "loss": 0.5781,
+      "step": 4848
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2116030433056135,
+      "learning_rate": 1.4346435875586324e-05,
+      "loss": 0.6044,
+      "step": 4849
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1843054114358138,
+      "learning_rate": 1.434417282042186e-05,
+      "loss": 0.5525,
+      "step": 4850
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1525653690658018,
+      "learning_rate": 1.434190949098735e-05,
+      "loss": 0.5536,
+      "step": 4851
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2550317104482467,
+      "learning_rate": 1.4339645887425693e-05,
+      "loss": 0.6192,
+      "step": 4852
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2371649779919311,
+      "learning_rate": 1.4337382009879806e-05,
+      "loss": 0.5949,
+      "step": 4853
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1263838839364222,
+      "learning_rate": 1.433511785849261e-05,
+      "loss": 0.5938,
+      "step": 4854
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.234956628447132,
+      "learning_rate": 1.433285343340706e-05,
+      "loss": 0.627,
+      "step": 4855
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0569472229484256,
+      "learning_rate": 1.4330588734766113e-05,
+      "loss": 0.5622,
+      "step": 4856
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2879857045262912,
+      "learning_rate": 1.432832376271276e-05,
+      "loss": 0.5922,
+      "step": 4857
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.107385759252496,
+      "learning_rate": 1.4326058517389998e-05,
+      "loss": 0.5996,
+      "step": 4858
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2098262487143545,
+      "learning_rate": 1.432379299894084e-05,
+      "loss": 0.5246,
+      "step": 4859
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.183172220133152,
+      "learning_rate": 1.4321527207508324e-05,
+      "loss": 0.6075,
+      "step": 4860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2062395579622367,
+      "learning_rate": 1.4319261143235496e-05,
+      "loss": 0.6014,
+      "step": 4861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2227280080779541,
+      "learning_rate": 1.431699480626543e-05,
+      "loss": 0.6177,
+      "step": 4862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1390181942291235,
+      "learning_rate": 1.431472819674121e-05,
+      "loss": 0.6039,
+      "step": 4863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2663982054580196,
+      "learning_rate": 1.4312461314805936e-05,
+      "loss": 0.6207,
+      "step": 4864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.099522588906349,
+      "learning_rate": 1.431019416060273e-05,
+      "loss": 0.5091,
+      "step": 4865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1178514868272384,
+      "learning_rate": 1.430792673427473e-05,
+      "loss": 0.5329,
+      "step": 4866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1970950529430149,
+      "learning_rate": 1.4305659035965087e-05,
+      "loss": 0.5721,
+      "step": 4867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1873983038472684,
+      "learning_rate": 1.430339106581698e-05,
+      "loss": 0.5644,
+      "step": 4868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2634657819980597,
+      "learning_rate": 1.4301122823973588e-05,
+      "loss": 0.5808,
+      "step": 4869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1288353392825703,
+      "learning_rate": 1.429885431057812e-05,
+      "loss": 0.5662,
+      "step": 4870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1901173608272502,
+      "learning_rate": 1.4296585525773803e-05,
+      "loss": 0.6045,
+      "step": 4871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1990183808872086,
+      "learning_rate": 1.4294316469703873e-05,
+      "loss": 0.5554,
+      "step": 4872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.162629954960051,
+      "learning_rate": 1.429204714251159e-05,
+      "loss": 0.5556,
+      "step": 4873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0618223946727976,
+      "learning_rate": 1.4289777544340227e-05,
+      "loss": 0.6083,
+      "step": 4874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2307555818030167,
+      "learning_rate": 1.4287507675333075e-05,
+      "loss": 0.6497,
+      "step": 4875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1254117530076173,
+      "learning_rate": 1.4285237535633442e-05,
+      "loss": 0.5986,
+      "step": 4876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1017090179182312,
+      "learning_rate": 1.4282967125384652e-05,
+      "loss": 0.5943,
+      "step": 4877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1277232059550661,
+      "learning_rate": 1.4280696444730057e-05,
+      "loss": 0.5673,
+      "step": 4878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.217047553776633,
+      "learning_rate": 1.4278425493813004e-05,
+      "loss": 0.6337,
+      "step": 4879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1508302739015464,
+      "learning_rate": 1.4276154272776876e-05,
+      "loss": 0.5765,
+      "step": 4880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1430070642501653,
+      "learning_rate": 1.4273882781765069e-05,
+      "loss": 0.5336,
+      "step": 4881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0953040390667381,
+      "learning_rate": 1.427161102092099e-05,
+      "loss": 0.5424,
+      "step": 4882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2456077177893994,
+      "learning_rate": 1.4269338990388072e-05,
+      "loss": 0.5926,
+      "step": 4883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.157212140601062,
+      "learning_rate": 1.4267066690309754e-05,
+      "loss": 0.5766,
+      "step": 4884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2686799936758892,
+      "learning_rate": 1.4264794120829499e-05,
+      "loss": 0.6392,
+      "step": 4885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3063366455106868,
+      "learning_rate": 1.4262521282090791e-05,
+      "loss": 0.6272,
+      "step": 4886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0878681064103706,
+      "learning_rate": 1.4260248174237121e-05,
+      "loss": 0.6112,
+      "step": 4887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1843401377838811,
+      "learning_rate": 1.4257974797412006e-05,
+      "loss": 0.6256,
+      "step": 4888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2296357256642696,
+      "learning_rate": 1.4255701151758972e-05,
+      "loss": 0.6256,
+      "step": 4889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0200358344309548,
+      "learning_rate": 1.4253427237421567e-05,
+      "loss": 0.5213,
+      "step": 4890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2829383797628846,
+      "learning_rate": 1.4251153054543357e-05,
+      "loss": 0.6187,
+      "step": 4891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0860861806834865,
+      "learning_rate": 1.4248878603267922e-05,
+      "loss": 0.5749,
+      "step": 4892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1488703735842798,
+      "learning_rate": 1.4246603883738859e-05,
+      "loss": 0.6043,
+      "step": 4893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1044169568142737,
+      "learning_rate": 1.4244328896099782e-05,
+      "loss": 0.5685,
+      "step": 4894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1651589778762321,
+      "learning_rate": 1.4242053640494322e-05,
+      "loss": 0.5881,
+      "step": 4895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1452661960291812,
+      "learning_rate": 1.4239778117066132e-05,
+      "loss": 0.5642,
+      "step": 4896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.048361903732733,
+      "learning_rate": 1.4237502325958876e-05,
+      "loss": 0.5737,
+      "step": 4897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2277367189186532,
+      "learning_rate": 1.4235226267316234e-05,
+      "loss": 0.5551,
+      "step": 4898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3070099313004027,
+      "learning_rate": 1.4232949941281908e-05,
+      "loss": 0.6612,
+      "step": 4899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2529759268430862,
+      "learning_rate": 1.4230673347999612e-05,
+      "loss": 0.627,
+      "step": 4900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2075407563907425,
+      "learning_rate": 1.4228396487613081e-05,
+      "loss": 0.6432,
+      "step": 4901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.125172972627046,
+      "learning_rate": 1.4226119360266065e-05,
+      "loss": 0.5594,
+      "step": 4902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3071368861676242,
+      "learning_rate": 1.4223841966102327e-05,
+      "loss": 0.6807,
+      "step": 4903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0771641526930442,
+      "learning_rate": 1.4221564305265657e-05,
+      "loss": 0.56,
+      "step": 4904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1093863365776122,
+      "learning_rate": 1.4219286377899848e-05,
+      "loss": 0.5597,
+      "step": 4905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1987146069629033,
+      "learning_rate": 1.4217008184148727e-05,
+      "loss": 0.5724,
+      "step": 4906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3141783473854975,
+      "learning_rate": 1.4214729724156118e-05,
+      "loss": 0.6102,
+      "step": 4907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2733609697245454,
+      "learning_rate": 1.4212450998065876e-05,
+      "loss": 0.6545,
+      "step": 4908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1138952870765229,
+      "learning_rate": 1.4210172006021872e-05,
+      "loss": 0.5782,
+      "step": 4909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0836727759843126,
+      "learning_rate": 1.4207892748167985e-05,
+      "loss": 0.5971,
+      "step": 4910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1116477359225532,
+      "learning_rate": 1.420561322464812e-05,
+      "loss": 0.5668,
+      "step": 4911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.188494466679667,
+      "learning_rate": 1.4203333435606196e-05,
+      "loss": 0.5782,
+      "step": 4912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.276618728426569,
+      "learning_rate": 1.4201053381186141e-05,
+      "loss": 0.5927,
+      "step": 4913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1233440503553442,
+      "learning_rate": 1.4198773061531917e-05,
+      "loss": 0.5638,
+      "step": 4914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9997992612106881,
+      "learning_rate": 1.4196492476787482e-05,
+      "loss": 0.5276,
+      "step": 4915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2454187843217115,
+      "learning_rate": 1.419421162709683e-05,
+      "loss": 0.6094,
+      "step": 4916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1780997701113802,
+      "learning_rate": 1.4191930512603956e-05,
+      "loss": 0.5681,
+      "step": 4917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.295072060645785,
+      "learning_rate": 1.4189649133452881e-05,
+      "loss": 0.5976,
+      "step": 4918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0245101200011653,
+      "learning_rate": 1.4187367489787642e-05,
+      "loss": 0.5122,
+      "step": 4919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2923113024446344,
+      "learning_rate": 1.4185085581752289e-05,
+      "loss": 0.5874,
+      "step": 4920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1116119719513675,
+      "learning_rate": 1.4182803409490891e-05,
+      "loss": 0.59,
+      "step": 4921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.260335346887879,
+      "learning_rate": 1.4180520973147534e-05,
+      "loss": 0.6102,
+      "step": 4922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2737362283068865,
+      "learning_rate": 1.4178238272866316e-05,
+      "loss": 0.5511,
+      "step": 4923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1566248105818988,
+      "learning_rate": 1.4175955308791363e-05,
+      "loss": 0.5751,
+      "step": 4924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2336226476240444,
+      "learning_rate": 1.4173672081066806e-05,
+      "loss": 0.569,
+      "step": 4925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2475597881751865,
+      "learning_rate": 1.4171388589836792e-05,
+      "loss": 0.5867,
+      "step": 4926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0598683584090418,
+      "learning_rate": 1.4169104835245502e-05,
+      "loss": 0.5513,
+      "step": 4927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2302143621122017,
+      "learning_rate": 1.416682081743711e-05,
+      "loss": 0.6471,
+      "step": 4928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.239611515825394,
+      "learning_rate": 1.4164536536555824e-05,
+      "loss": 0.5612,
+      "step": 4929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1506194063679558,
+      "learning_rate": 1.4162251992745858e-05,
+      "loss": 0.5476,
+      "step": 4930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1681051015205106,
+      "learning_rate": 1.4159967186151446e-05,
+      "loss": 0.6005,
+      "step": 4931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3163056377595899,
+      "learning_rate": 1.4157682116916851e-05,
+      "loss": 0.6112,
+      "step": 4932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.256711584666228,
+      "learning_rate": 1.415539678518633e-05,
+      "loss": 0.6149,
+      "step": 4933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.202299007899473,
+      "learning_rate": 1.415311119110417e-05,
+      "loss": 0.6107,
+      "step": 4934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3072290392866028,
+      "learning_rate": 1.415082533481467e-05,
+      "loss": 0.6436,
+      "step": 4935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1667086843462804,
+      "learning_rate": 1.4148539216462153e-05,
+      "loss": 0.6443,
+      "step": 4936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1747074229628005,
+      "learning_rate": 1.4146252836190958e-05,
+      "loss": 0.5495,
+      "step": 4937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2530183589774615,
+      "learning_rate": 1.4143966194145424e-05,
+      "loss": 0.6058,
+      "step": 4938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0756300654449973,
+      "learning_rate": 1.4141679290469925e-05,
+      "loss": 0.5889,
+      "step": 4939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.251629482577088,
+      "learning_rate": 1.4139392125308842e-05,
+      "loss": 0.643,
+      "step": 4940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1757323803044788,
+      "learning_rate": 1.413710469880658e-05,
+      "loss": 0.5421,
+      "step": 4941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2961164807056123,
+      "learning_rate": 1.4134817011107555e-05,
+      "loss": 0.5945,
+      "step": 4942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.155718913411746,
+      "learning_rate": 1.4132529062356197e-05,
+      "loss": 0.5825,
+      "step": 4943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1206681776265388,
+      "learning_rate": 1.4130240852696958e-05,
+      "loss": 0.5698,
+      "step": 4944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0634337417486805,
+      "learning_rate": 1.4127952382274305e-05,
+      "loss": 0.5843,
+      "step": 4945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1516916790043479,
+      "learning_rate": 1.4125663651232725e-05,
+      "loss": 0.5607,
+      "step": 4946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.188723084860073,
+      "learning_rate": 1.4123374659716708e-05,
+      "loss": 0.5515,
+      "step": 4947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1545604527515752,
+      "learning_rate": 1.4121085407870776e-05,
+      "loss": 0.5645,
+      "step": 4948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.04304921272021,
+      "learning_rate": 1.411879589583946e-05,
+      "loss": 0.5281,
+      "step": 4949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.142491942195733,
+      "learning_rate": 1.4116506123767308e-05,
+      "loss": 0.5592,
+      "step": 4950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1670457133383783,
+      "learning_rate": 1.411421609179889e-05,
+      "loss": 0.5679,
+      "step": 4951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1221702272237901,
+      "learning_rate": 1.411192580007878e-05,
+      "loss": 0.5746,
+      "step": 4952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.189275167648375,
+      "learning_rate": 1.410963524875158e-05,
+      "loss": 0.5174,
+      "step": 4953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2529479075271828,
+      "learning_rate": 1.4107344437961902e-05,
+      "loss": 0.5678,
+      "step": 4954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2215801536941462,
+      "learning_rate": 1.4105053367854382e-05,
+      "loss": 0.6142,
+      "step": 4955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1335327489570246,
+      "learning_rate": 1.4102762038573663e-05,
+      "loss": 0.5655,
+      "step": 4956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2208438883684272,
+      "learning_rate": 1.410047045026441e-05,
+      "loss": 0.6312,
+      "step": 4957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.177590736246212,
+      "learning_rate": 1.40981786030713e-05,
+      "loss": 0.5922,
+      "step": 4958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.097529556727991,
+      "learning_rate": 1.4095886497139029e-05,
+      "loss": 0.5457,
+      "step": 4959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1193182306681342,
+      "learning_rate": 1.4093594132612314e-05,
+      "loss": 0.4991,
+      "step": 4960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1690475941767988,
+      "learning_rate": 1.4091301509635882e-05,
+      "loss": 0.533,
+      "step": 4961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.168178067540347,
+      "learning_rate": 1.4089008628354477e-05,
+      "loss": 0.6165,
+      "step": 4962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1656709861377175,
+      "learning_rate": 1.4086715488912858e-05,
+      "loss": 0.6242,
+      "step": 4963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2235963272537176,
+      "learning_rate": 1.4084422091455808e-05,
+      "loss": 0.6116,
+      "step": 4964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2199280254051401,
+      "learning_rate": 1.408212843612812e-05,
+      "loss": 0.5957,
+      "step": 4965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.090983625325818,
+      "learning_rate": 1.40798345230746e-05,
+      "loss": 0.571,
+      "step": 4966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1602652142149223,
+      "learning_rate": 1.407754035244008e-05,
+      "loss": 0.5741,
+      "step": 4967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1282554783484182,
+      "learning_rate": 1.4075245924369397e-05,
+      "loss": 0.5207,
+      "step": 4968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1144929950266353,
+      "learning_rate": 1.4072951239007414e-05,
+      "loss": 0.5441,
+      "step": 4969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2034443518147127,
+      "learning_rate": 1.4070656296499006e-05,
+      "loss": 0.6181,
+      "step": 4970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2046234410691394,
+      "learning_rate": 1.4068361096989064e-05,
+      "loss": 0.5711,
+      "step": 4971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.270449403186937,
+      "learning_rate": 1.4066065640622499e-05,
+      "loss": 0.6,
+      "step": 4972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1815090137675432,
+      "learning_rate": 1.4063769927544228e-05,
+      "loss": 0.5679,
+      "step": 4973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1709912846319916,
+      "learning_rate": 1.4061473957899194e-05,
+      "loss": 0.5673,
+      "step": 4974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.194542281384923,
+      "learning_rate": 1.4059177731832358e-05,
+      "loss": 0.6002,
+      "step": 4975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1589470929666976,
+      "learning_rate": 1.4056881249488692e-05,
+      "loss": 0.5775,
+      "step": 4976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1443728184327646,
+      "learning_rate": 1.4054584511013178e-05,
+      "loss": 0.5174,
+      "step": 4977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.202087053704193,
+      "learning_rate": 1.4052287516550824e-05,
+      "loss": 0.5972,
+      "step": 4978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2615736650601708,
+      "learning_rate": 1.4049990266246652e-05,
+      "loss": 0.6203,
+      "step": 4979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.174427761579736,
+      "learning_rate": 1.4047692760245702e-05,
+      "loss": 0.5885,
+      "step": 4980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1007120623751325,
+      "learning_rate": 1.4045394998693022e-05,
+      "loss": 0.5585,
+      "step": 4981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2990526048376556,
+      "learning_rate": 1.4043096981733687e-05,
+      "loss": 0.6313,
+      "step": 4982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.195364857443232,
+      "learning_rate": 1.4040798709512777e-05,
+      "loss": 0.5521,
+      "step": 4983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.148793327585384,
+      "learning_rate": 1.4038500182175399e-05,
+      "loss": 0.5952,
+      "step": 4984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2284884548366712,
+      "learning_rate": 1.4036201399866669e-05,
+      "loss": 0.5935,
+      "step": 4985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.197719817642587,
+      "learning_rate": 1.4033902362731719e-05,
+      "loss": 0.5765,
+      "step": 4986
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1187142030753594,
+      "learning_rate": 1.40316030709157e-05,
+      "loss": 0.5516,
+      "step": 4987
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1827435351162627,
+      "learning_rate": 1.402930352456378e-05,
+      "loss": 0.5638,
+      "step": 4988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2784737537869166,
+      "learning_rate": 1.402700372382114e-05,
+      "loss": 0.5931,
+      "step": 4989
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1319007624575315,
+      "learning_rate": 1.4024703668832978e-05,
+      "loss": 0.5451,
+      "step": 4990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2729598477662063,
+      "learning_rate": 1.402240335974451e-05,
+      "loss": 0.5577,
+      "step": 4991
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2129385381255733,
+      "learning_rate": 1.4020102796700962e-05,
+      "loss": 0.5971,
+      "step": 4992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1671036287990175,
+      "learning_rate": 1.4017801979847586e-05,
+      "loss": 0.6157,
+      "step": 4993
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1993560712393094,
+      "learning_rate": 1.401550090932964e-05,
+      "loss": 0.5694,
+      "step": 4994
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1787565992055329,
+      "learning_rate": 1.4013199585292405e-05,
+      "loss": 0.5989,
+      "step": 4995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1017810253070914,
+      "learning_rate": 1.4010898007881177e-05,
+      "loss": 0.5756,
+      "step": 4996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0318563008361465,
+      "learning_rate": 1.400859617724126e-05,
+      "loss": 0.5273,
+      "step": 4997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0754525607086431,
+      "learning_rate": 1.4006294093517989e-05,
+      "loss": 0.5224,
+      "step": 4998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0694144720911443,
+      "learning_rate": 1.4003991756856699e-05,
+      "loss": 0.5834,
+      "step": 4999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.250856392275336,
+      "learning_rate": 1.4001689167402752e-05,
+      "loss": 0.5918,
+      "step": 5000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1297402122936293,
+      "learning_rate": 1.3999386325301525e-05,
+      "loss": 0.6011,
+      "step": 5001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.10318884948437,
+      "learning_rate": 1.39970832306984e-05,
+      "loss": 0.5746,
+      "step": 5002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1938638822548917,
+      "learning_rate": 1.3994779883738794e-05,
+      "loss": 0.6109,
+      "step": 5003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1559941807834713,
+      "learning_rate": 1.399247628456812e-05,
+      "loss": 0.5714,
+      "step": 5004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1841538114518846,
+      "learning_rate": 1.3990172433331819e-05,
+      "loss": 0.5876,
+      "step": 5005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.182314745079797,
+      "learning_rate": 1.398786833017535e-05,
+      "loss": 0.5583,
+      "step": 5006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0280871584343054,
+      "learning_rate": 1.3985563975244175e-05,
+      "loss": 0.536,
+      "step": 5007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1811960944738003,
+      "learning_rate": 1.3983259368683784e-05,
+      "loss": 0.5858,
+      "step": 5008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1201477312681183,
+      "learning_rate": 1.3980954510639677e-05,
+      "loss": 0.5802,
+      "step": 5009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.16789270874303,
+      "learning_rate": 1.3978649401257375e-05,
+      "loss": 0.6132,
+      "step": 5010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.271520611177799,
+      "learning_rate": 1.3976344040682409e-05,
+      "loss": 0.6627,
+      "step": 5011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1646887099758994,
+      "learning_rate": 1.3974038429060326e-05,
+      "loss": 0.5353,
+      "step": 5012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0602171439755919,
+      "learning_rate": 1.3971732566536695e-05,
+      "loss": 0.526,
+      "step": 5013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1034120224112942,
+      "learning_rate": 1.3969426453257096e-05,
+      "loss": 0.5937,
+      "step": 5014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2315280290574246,
+      "learning_rate": 1.3967120089367128e-05,
+      "loss": 0.5864,
+      "step": 5015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0512289915582456,
+      "learning_rate": 1.3964813475012398e-05,
+      "loss": 0.5216,
+      "step": 5016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.198551750456754,
+      "learning_rate": 1.3962506610338537e-05,
+      "loss": 0.5415,
+      "step": 5017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.160523173559806,
+      "learning_rate": 1.3960199495491192e-05,
+      "loss": 0.5936,
+      "step": 5018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0838572934551802,
+      "learning_rate": 1.3957892130616018e-05,
+      "loss": 0.5861,
+      "step": 5019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2284209634914356,
+      "learning_rate": 1.3955584515858696e-05,
+      "loss": 0.6109,
+      "step": 5020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1760741721234578,
+      "learning_rate": 1.3953276651364918e-05,
+      "loss": 0.5597,
+      "step": 5021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2884355812326782,
+      "learning_rate": 1.3950968537280383e-05,
+      "loss": 0.5995,
+      "step": 5022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1295864603488763,
+      "learning_rate": 1.3948660173750823e-05,
+      "loss": 0.5355,
+      "step": 5023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1063434927706548,
+      "learning_rate": 1.3946351560921974e-05,
+      "loss": 0.5495,
+      "step": 5024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2446574001709683,
+      "learning_rate": 1.3944042698939591e-05,
+      "loss": 0.6116,
+      "step": 5025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1797739464620633,
+      "learning_rate": 1.3941733587949442e-05,
+      "loss": 0.6027,
+      "step": 5026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1064585645375302,
+      "learning_rate": 1.3939424228097314e-05,
+      "loss": 0.5717,
+      "step": 5027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1813293551477768,
+      "learning_rate": 1.3937114619529012e-05,
+      "loss": 0.5832,
+      "step": 5028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.114436838142661,
+      "learning_rate": 1.3934804762390351e-05,
+      "loss": 0.5662,
+      "step": 5029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1960966741823895,
+      "learning_rate": 1.3932494656827165e-05,
+      "loss": 0.5762,
+      "step": 5030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1692603888891244,
+      "learning_rate": 1.3930184302985302e-05,
+      "loss": 0.5788,
+      "step": 5031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1236667680309351,
+      "learning_rate": 1.3927873701010625e-05,
+      "loss": 0.5405,
+      "step": 5032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2086853742237078,
+      "learning_rate": 1.3925562851049017e-05,
+      "loss": 0.6183,
+      "step": 5033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1783311132349865,
+      "learning_rate": 1.3923251753246371e-05,
+      "loss": 0.5797,
+      "step": 5034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2271953820892694,
+      "learning_rate": 1.3920940407748607e-05,
+      "loss": 0.6282,
+      "step": 5035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2846407798636061,
+      "learning_rate": 1.3918628814701643e-05,
+      "loss": 0.6647,
+      "step": 5036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1136105234513498,
+      "learning_rate": 1.391631697425142e-05,
+      "loss": 0.5859,
+      "step": 5037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1289616871458,
+      "learning_rate": 1.3914004886543904e-05,
+      "loss": 0.586,
+      "step": 5038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.097468512828539,
+      "learning_rate": 1.3911692551725065e-05,
+      "loss": 0.4819,
+      "step": 5039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2165046941162556,
+      "learning_rate": 1.3909379969940894e-05,
+      "loss": 0.516,
+      "step": 5040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.192142846707376,
+      "learning_rate": 1.3907067141337397e-05,
+      "loss": 0.5959,
+      "step": 5041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0752386072296287,
+      "learning_rate": 1.3904754066060593e-05,
+      "loss": 0.5593,
+      "step": 5042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1220058226748795,
+      "learning_rate": 1.3902440744256516e-05,
+      "loss": 0.5524,
+      "step": 5043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0781451237224033,
+      "learning_rate": 1.3900127176071226e-05,
+      "loss": 0.5714,
+      "step": 5044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1659100099715796,
+      "learning_rate": 1.3897813361650783e-05,
+      "loss": 0.5683,
+      "step": 5045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1880990073598108,
+      "learning_rate": 1.3895499301141273e-05,
+      "loss": 0.5649,
+      "step": 5046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.134200935565453,
+      "learning_rate": 1.389318499468879e-05,
+      "loss": 0.59,
+      "step": 5047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2836453910929442,
+      "learning_rate": 1.3890870442439455e-05,
+      "loss": 0.559,
+      "step": 5048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1648478578921058,
+      "learning_rate": 1.3888555644539397e-05,
+      "loss": 0.6468,
+      "step": 5049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1400754466940612,
+      "learning_rate": 1.3886240601134757e-05,
+      "loss": 0.5487,
+      "step": 5050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0193843808844685,
+      "learning_rate": 1.3883925312371698e-05,
+      "loss": 0.5642,
+      "step": 5051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0672951930207055,
+      "learning_rate": 1.3881609778396394e-05,
+      "loss": 0.5826,
+      "step": 5052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.172555192002202,
+      "learning_rate": 1.3879293999355037e-05,
+      "loss": 0.5745,
+      "step": 5053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1489998096070826,
+      "learning_rate": 1.387697797539384e-05,
+      "loss": 0.5357,
+      "step": 5054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2987461150242523,
+      "learning_rate": 1.3874661706659018e-05,
+      "loss": 0.5867,
+      "step": 5055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1839521010382115,
+      "learning_rate": 1.3872345193296812e-05,
+      "loss": 0.6301,
+      "step": 5056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0940891285074374,
+      "learning_rate": 1.3870028435453476e-05,
+      "loss": 0.5202,
+      "step": 5057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1252981426633528,
+      "learning_rate": 1.3867711433275275e-05,
+      "loss": 0.6074,
+      "step": 5058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1649098224693075,
+      "learning_rate": 1.38653941869085e-05,
+      "loss": 0.5616,
+      "step": 5059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1961083349613146,
+      "learning_rate": 1.3863076696499447e-05,
+      "loss": 0.5461,
+      "step": 5060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0949045354781428,
+      "learning_rate": 1.3860758962194432e-05,
+      "loss": 0.5873,
+      "step": 5061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3233252918005483,
+      "learning_rate": 1.385844098413978e-05,
+      "loss": 0.605,
+      "step": 5062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.225336081276207,
+      "learning_rate": 1.3856122762481845e-05,
+      "loss": 0.5775,
+      "step": 5063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.189911251445215,
+      "learning_rate": 1.3853804297366986e-05,
+      "loss": 0.5683,
+      "step": 5064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1688029904515929,
+      "learning_rate": 1.3851485588941578e-05,
+      "loss": 0.569,
+      "step": 5065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1842199500863406,
+      "learning_rate": 1.3849166637352011e-05,
+      "loss": 0.5554,
+      "step": 5066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1851311697974825,
+      "learning_rate": 1.38468474427447e-05,
+      "loss": 0.5569,
+      "step": 5067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0676355228564274,
+      "learning_rate": 1.3844528005266057e-05,
+      "loss": 0.5444,
+      "step": 5068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0443988953007817,
+      "learning_rate": 1.3842208325062532e-05,
+      "loss": 0.5232,
+      "step": 5069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1501182142948272,
+      "learning_rate": 1.383988840228057e-05,
+      "loss": 0.553,
+      "step": 5070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0773497848950297,
+      "learning_rate": 1.3837568237066641e-05,
+      "loss": 0.5317,
+      "step": 5071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0995766887244798,
+      "learning_rate": 1.383524782956723e-05,
+      "loss": 0.5444,
+      "step": 5072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1433419046738715,
+      "learning_rate": 1.3832927179928834e-05,
+      "loss": 0.5872,
+      "step": 5073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2043275644238283,
+      "learning_rate": 1.3830606288297976e-05,
+      "loss": 0.5952,
+      "step": 5074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2522080945881557,
+      "learning_rate": 1.3828285154821175e-05,
+      "loss": 0.5521,
+      "step": 5075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1262906935746209,
+      "learning_rate": 1.3825963779644981e-05,
+      "loss": 0.575,
+      "step": 5076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1679750780425355,
+      "learning_rate": 1.3823642162915958e-05,
+      "loss": 0.4945,
+      "step": 5077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2046123575210388,
+      "learning_rate": 1.3821320304780671e-05,
+      "loss": 0.5832,
+      "step": 5078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0860870038864228,
+      "learning_rate": 1.3818998205385726e-05,
+      "loss": 0.512,
+      "step": 5079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0958168664330807,
+      "learning_rate": 1.3816675864877715e-05,
+      "loss": 0.5357,
+      "step": 5080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.14363766311371,
+      "learning_rate": 1.3814353283403266e-05,
+      "loss": 0.5657,
+      "step": 5081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0858749261844856,
+      "learning_rate": 1.3812030461109017e-05,
+      "loss": 0.4903,
+      "step": 5082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2557622180809869,
+      "learning_rate": 1.3809707398141615e-05,
+      "loss": 0.6093,
+      "step": 5083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.277200346663464,
+      "learning_rate": 1.3807384094647728e-05,
+      "loss": 0.6032,
+      "step": 5084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2659834836783417,
+      "learning_rate": 1.3805060550774045e-05,
+      "loss": 0.5976,
+      "step": 5085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1807909219771489,
+      "learning_rate": 1.3802736766667253e-05,
+      "loss": 0.5401,
+      "step": 5086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2161290768674509,
+      "learning_rate": 1.3800412742474074e-05,
+      "loss": 0.6131,
+      "step": 5087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2514854664615915,
+      "learning_rate": 1.3798088478341226e-05,
+      "loss": 0.6247,
+      "step": 5088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2108117315138005,
+      "learning_rate": 1.3795763974415459e-05,
+      "loss": 0.5639,
+      "step": 5089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.171089417676693,
+      "learning_rate": 1.379343923084353e-05,
+      "loss": 0.5985,
+      "step": 5090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0987285114569603,
+      "learning_rate": 1.379111424777221e-05,
+      "loss": 0.5674,
+      "step": 5091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2470046394280194,
+      "learning_rate": 1.3788789025348288e-05,
+      "loss": 0.5939,
+      "step": 5092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2564369404183413,
+      "learning_rate": 1.3786463563718564e-05,
+      "loss": 0.6035,
+      "step": 5093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3448791638942845,
+      "learning_rate": 1.3784137863029864e-05,
+      "loss": 0.6392,
+      "step": 5094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3296999848941775,
+      "learning_rate": 1.3781811923429018e-05,
+      "loss": 0.6535,
+      "step": 5095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1156698201897324,
+      "learning_rate": 1.3779485745062869e-05,
+      "loss": 0.5705,
+      "step": 5096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1670516888760163,
+      "learning_rate": 1.3777159328078289e-05,
+      "loss": 0.621,
+      "step": 5097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1223376879341824,
+      "learning_rate": 1.377483267262215e-05,
+      "loss": 0.5483,
+      "step": 5098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0578393817230047,
+      "learning_rate": 1.3772505778841351e-05,
+      "loss": 0.532,
+      "step": 5099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1010552719752902,
+      "learning_rate": 1.37701786468828e-05,
+      "loss": 0.5705,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2344446283331352,
+      "learning_rate": 1.3767851276893415e-05,
+      "loss": 0.5687,
+      "step": 5101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1179940573375722,
+      "learning_rate": 1.3765523669020144e-05,
+      "loss": 0.528,
+      "step": 5102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2326976147245252,
+      "learning_rate": 1.3763195823409932e-05,
+      "loss": 0.5715,
+      "step": 5103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1652079841062606,
+      "learning_rate": 1.3760867740209755e-05,
+      "loss": 0.5853,
+      "step": 5104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.096963303970221,
+      "learning_rate": 1.3758539419566595e-05,
+      "loss": 0.5749,
+      "step": 5105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0746071408136593,
+      "learning_rate": 1.3756210861627446e-05,
+      "loss": 0.5525,
+      "step": 5106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0965460324729155,
+      "learning_rate": 1.3753882066539328e-05,
+      "loss": 0.5335,
+      "step": 5107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1645182287116982,
+      "learning_rate": 1.3751553034449267e-05,
+      "loss": 0.6002,
+      "step": 5108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.157325038663377,
+      "learning_rate": 1.374922376550431e-05,
+      "loss": 0.5593,
+      "step": 5109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1870787023791176,
+      "learning_rate": 1.3746894259851513e-05,
+      "loss": 0.5456,
+      "step": 5110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.302207233891866,
+      "learning_rate": 1.3744564517637947e-05,
+      "loss": 0.5702,
+      "step": 5111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1542231863805963,
+      "learning_rate": 1.3742234539010703e-05,
+      "loss": 0.5469,
+      "step": 5112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1879417701957935,
+      "learning_rate": 1.3739904324116888e-05,
+      "loss": 0.5793,
+      "step": 5113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1136928935026065,
+      "learning_rate": 1.3737573873103615e-05,
+      "loss": 0.5757,
+      "step": 5114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2274395182328943,
+      "learning_rate": 1.3735243186118026e-05,
+      "loss": 0.5635,
+      "step": 5115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1150440915546074,
+      "learning_rate": 1.3732912263307256e-05,
+      "loss": 0.5287,
+      "step": 5116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.225364099616156,
+      "learning_rate": 1.3730581104818477e-05,
+      "loss": 0.5333,
+      "step": 5117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1317513066430256,
+      "learning_rate": 1.372824971079887e-05,
+      "loss": 0.5736,
+      "step": 5118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.209555988420983,
+      "learning_rate": 1.3725918081395619e-05,
+      "loss": 0.5854,
+      "step": 5119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1674513675599836,
+      "learning_rate": 1.372358621675594e-05,
+      "loss": 0.5563,
+      "step": 5120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.223065390091859,
+      "learning_rate": 1.3721254117027049e-05,
+      "loss": 0.6077,
+      "step": 5121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1131230827542071,
+      "learning_rate": 1.3718921782356187e-05,
+      "loss": 0.5505,
+      "step": 5122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.270391835718321,
+      "learning_rate": 1.3716589212890608e-05,
+      "loss": 0.6009,
+      "step": 5123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2485479026743516,
+      "learning_rate": 1.3714256408777575e-05,
+      "loss": 0.6012,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.208180445283391,
+      "learning_rate": 1.3711923370164373e-05,
+      "loss": 0.6001,
+      "step": 5125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2076025046206034,
+      "learning_rate": 1.3709590097198295e-05,
+      "loss": 0.5781,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0729987412451536,
+      "learning_rate": 1.3707256590026656e-05,
+      "loss": 0.517,
+      "step": 5127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.181185144329341,
+      "learning_rate": 1.3704922848796782e-05,
+      "loss": 0.5925,
+      "step": 5128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.14652955260288,
+      "learning_rate": 1.3702588873656015e-05,
+      "loss": 0.5673,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2077652264383187,
+      "learning_rate": 1.370025466475171e-05,
+      "loss": 0.6318,
+      "step": 5130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2391190447418405,
+      "learning_rate": 1.3697920222231233e-05,
+      "loss": 0.5819,
+      "step": 5131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0633947868187816,
+      "learning_rate": 1.3695585546241977e-05,
+      "loss": 0.5433,
+      "step": 5132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1979733937418402,
+      "learning_rate": 1.3693250636931339e-05,
+      "loss": 0.571,
+      "step": 5133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.278536318603655,
+      "learning_rate": 1.3690915494446732e-05,
+      "loss": 0.5792,
+      "step": 5134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2121630957620921,
+      "learning_rate": 1.3688580118935591e-05,
+      "loss": 0.5875,
+      "step": 5135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1831097510272532,
+      "learning_rate": 1.3686244510545353e-05,
+      "loss": 0.5541,
+      "step": 5136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1002999503653201,
+      "learning_rate": 1.3683908669423479e-05,
+      "loss": 0.5572,
+      "step": 5137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1298810188984791,
+      "learning_rate": 1.3681572595717446e-05,
+      "loss": 0.584,
+      "step": 5138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0699068400568583,
+      "learning_rate": 1.3679236289574742e-05,
+      "loss": 0.5415,
+      "step": 5139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1055363610825557,
+      "learning_rate": 1.3676899751142866e-05,
+      "loss": 0.5655,
+      "step": 5140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1047699677630374,
+      "learning_rate": 1.3674562980569342e-05,
+      "loss": 0.6065,
+      "step": 5141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1178692424855816,
+      "learning_rate": 1.3672225978001694e-05,
+      "loss": 0.5032,
+      "step": 5142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1723398939514398,
+      "learning_rate": 1.3669888743587478e-05,
+      "loss": 0.6144,
+      "step": 5143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1890604407634775,
+      "learning_rate": 1.3667551277474252e-05,
+      "loss": 0.5509,
+      "step": 5144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2264492687272461,
+      "learning_rate": 1.3665213579809588e-05,
+      "loss": 0.6048,
+      "step": 5145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1425296088457548,
+      "learning_rate": 1.3662875650741082e-05,
+      "loss": 0.558,
+      "step": 5146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1903742588773563,
+      "learning_rate": 1.3660537490416337e-05,
+      "loss": 0.6399,
+      "step": 5147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.059610364306384,
+      "learning_rate": 1.3658199098982979e-05,
+      "loss": 0.5719,
+      "step": 5148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3932991034046338,
+      "learning_rate": 1.3655860476588636e-05,
+      "loss": 0.6588,
+      "step": 5149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3501578733180393,
+      "learning_rate": 1.3653521623380956e-05,
+      "loss": 0.6182,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.133479691386181,
+      "learning_rate": 1.365118253950761e-05,
+      "loss": 0.5656,
+      "step": 5151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1529509932565465,
+      "learning_rate": 1.364884322511627e-05,
+      "loss": 0.5785,
+      "step": 5152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1952192980763319,
+      "learning_rate": 1.3646503680354633e-05,
+      "loss": 0.5223,
+      "step": 5153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0270377163819115,
+      "learning_rate": 1.3644163905370406e-05,
+      "loss": 0.5162,
+      "step": 5154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.081539276323409,
+      "learning_rate": 1.3641823900311308e-05,
+      "loss": 0.4914,
+      "step": 5155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2481224741597374,
+      "learning_rate": 1.3639483665325082e-05,
+      "loss": 0.5705,
+      "step": 5156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.239927671461129,
+      "learning_rate": 1.363714320055947e-05,
+      "loss": 0.6092,
+      "step": 5157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1684606020434887,
+      "learning_rate": 1.3634802506162246e-05,
+      "loss": 0.5975,
+      "step": 5158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.159317484043346,
+      "learning_rate": 1.3632461582281184e-05,
+      "loss": 0.536,
+      "step": 5159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2963703966280629,
+      "learning_rate": 1.3630120429064082e-05,
+      "loss": 0.6333,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1223450698642972,
+      "learning_rate": 1.3627779046658752e-05,
+      "loss": 0.5367,
+      "step": 5161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.152854780257862,
+      "learning_rate": 1.3625437435213009e-05,
+      "loss": 0.5699,
+      "step": 5162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1573887964800986,
+      "learning_rate": 1.3623095594874698e-05,
+      "loss": 0.5517,
+      "step": 5163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1705242255766188,
+      "learning_rate": 1.362075352579167e-05,
+      "loss": 0.5387,
+      "step": 5164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2282702954566522,
+      "learning_rate": 1.361841122811179e-05,
+      "loss": 0.5727,
+      "step": 5165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1651345251451357,
+      "learning_rate": 1.361606870198294e-05,
+      "loss": 0.6193,
+      "step": 5166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2741724245595392,
+      "learning_rate": 1.3613725947553018e-05,
+      "loss": 0.5958,
+      "step": 5167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1994403547378543,
+      "learning_rate": 1.3611382964969933e-05,
+      "loss": 0.6221,
+      "step": 5168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2577098543921341,
+      "learning_rate": 1.360903975438161e-05,
+      "loss": 0.5529,
+      "step": 5169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2031053318546336,
+      "learning_rate": 1.3606696315935983e-05,
+      "loss": 0.5918,
+      "step": 5170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2037690655712658,
+      "learning_rate": 1.3604352649781012e-05,
+      "loss": 0.5967,
+      "step": 5171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1690285253903214,
+      "learning_rate": 1.3602008756064661e-05,
+      "loss": 0.6055,
+      "step": 5172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1573498108979243,
+      "learning_rate": 1.3599664634934918e-05,
+      "loss": 0.566,
+      "step": 5173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1630579050627143,
+      "learning_rate": 1.359732028653977e-05,
+      "loss": 0.5949,
+      "step": 5174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1540691841814061,
+      "learning_rate": 1.3594975711027231e-05,
+      "loss": 0.596,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1927771513740384,
+      "learning_rate": 1.3592630908545332e-05,
+      "loss": 0.5648,
+      "step": 5176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0961138113718683,
+      "learning_rate": 1.3590285879242107e-05,
+      "loss": 0.5902,
+      "step": 5177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1605104361725986,
+      "learning_rate": 1.358794062326561e-05,
+      "loss": 0.5967,
+      "step": 5178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0735656972680843,
+      "learning_rate": 1.3585595140763912e-05,
+      "loss": 0.5982,
+      "step": 5179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1756079156294366,
+      "learning_rate": 1.3583249431885091e-05,
+      "loss": 0.6237,
+      "step": 5180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.137341253500891,
+      "learning_rate": 1.3580903496777248e-05,
+      "loss": 0.5724,
+      "step": 5181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2936730440259925,
+      "learning_rate": 1.3578557335588492e-05,
+      "loss": 0.6853,
+      "step": 5182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.160855528862785,
+      "learning_rate": 1.3576210948466948e-05,
+      "loss": 0.5354,
+      "step": 5183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1667562972561536,
+      "learning_rate": 1.3573864335560756e-05,
+      "loss": 0.5645,
+      "step": 5184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1011772293658901,
+      "learning_rate": 1.357151749701807e-05,
+      "loss": 0.5405,
+      "step": 5185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1593123426754495,
+      "learning_rate": 1.3569170432987058e-05,
+      "loss": 0.5793,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.141263025633567,
+      "learning_rate": 1.35668231436159e-05,
+      "loss": 0.5594,
+      "step": 5187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1514834020977904,
+      "learning_rate": 1.3564475629052798e-05,
+      "loss": 0.5516,
+      "step": 5188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1484035045756322,
+      "learning_rate": 1.3562127889445958e-05,
+      "loss": 0.6041,
+      "step": 5189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.286528947053089,
+      "learning_rate": 1.3559779924943603e-05,
+      "loss": 0.6325,
+      "step": 5190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3521997790553795,
+      "learning_rate": 1.355743173569398e-05,
+      "loss": 0.6333,
+      "step": 5191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2142113514577728,
+      "learning_rate": 1.3555083321845335e-05,
+      "loss": 0.5448,
+      "step": 5192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1258377029273912,
+      "learning_rate": 1.355273468354594e-05,
+      "loss": 0.5584,
+      "step": 5193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1216301781064957,
+      "learning_rate": 1.3550385820944074e-05,
+      "loss": 0.5458,
+      "step": 5194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2637951184980942,
+      "learning_rate": 1.3548036734188033e-05,
+      "loss": 0.587,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2355631650851207,
+      "learning_rate": 1.3545687423426128e-05,
+      "loss": 0.6126,
+      "step": 5196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.085701951202441,
+      "learning_rate": 1.354333788880668e-05,
+      "loss": 0.4982,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1498156731535774,
+      "learning_rate": 1.3540988130478034e-05,
+      "loss": 0.5754,
+      "step": 5198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1721203356472376,
+      "learning_rate": 1.3538638148588537e-05,
+      "loss": 0.5825,
+      "step": 5199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2816878710325035,
+      "learning_rate": 1.3536287943286555e-05,
+      "loss": 0.5741,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2869727564360933,
+      "learning_rate": 1.353393751472047e-05,
+      "loss": 0.6197,
+      "step": 5201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1260877754494374,
+      "learning_rate": 1.3531586863038678e-05,
+      "loss": 0.5892,
+      "step": 5202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2899028004392399,
+      "learning_rate": 1.3529235988389588e-05,
+      "loss": 0.5654,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1726781000570679,
+      "learning_rate": 1.3526884890921621e-05,
+      "loss": 0.6146,
+      "step": 5204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2676549103554768,
+      "learning_rate": 1.3524533570783214e-05,
+      "loss": 0.6235,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1817948674189769,
+      "learning_rate": 1.3522182028122817e-05,
+      "loss": 0.5903,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1638322128774148,
+      "learning_rate": 1.3519830263088899e-05,
+      "loss": 0.555,
+      "step": 5207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7848086362996562,
+      "learning_rate": 1.3517478275829935e-05,
+      "loss": 0.5924,
+      "step": 5208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1323558018315305,
+      "learning_rate": 1.3515126066494422e-05,
+      "loss": 0.5676,
+      "step": 5209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.297026889592874,
+      "learning_rate": 1.351277363523086e-05,
+      "loss": 0.5744,
+      "step": 5210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2595048974681344,
+      "learning_rate": 1.3510420982187777e-05,
+      "loss": 0.6017,
+      "step": 5211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.185826527827758,
+      "learning_rate": 1.350806810751371e-05,
+      "loss": 0.5712,
+      "step": 5212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.202157957161739,
+      "learning_rate": 1.3505715011357202e-05,
+      "loss": 0.6262,
+      "step": 5213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.173455253700427,
+      "learning_rate": 1.3503361693866817e-05,
+      "loss": 0.5991,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1796111025293248,
+      "learning_rate": 1.350100815519113e-05,
+      "loss": 0.562,
+      "step": 5215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2338232907959152,
+      "learning_rate": 1.349865439547874e-05,
+      "loss": 0.6106,
+      "step": 5216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1161664551743316,
+      "learning_rate": 1.3496300414878247e-05,
+      "loss": 0.5736,
+      "step": 5217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2048698752332165,
+      "learning_rate": 1.3493946213538269e-05,
+      "loss": 0.5271,
+      "step": 5218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.165472981048526,
+      "learning_rate": 1.3491591791607443e-05,
+      "loss": 0.5496,
+      "step": 5219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1799031243725278,
+      "learning_rate": 1.3489237149234409e-05,
+      "loss": 0.5648,
+      "step": 5220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.318597435826858,
+      "learning_rate": 1.3486882286567831e-05,
+      "loss": 0.563,
+      "step": 5221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2489523789128933,
+      "learning_rate": 1.3484527203756388e-05,
+      "loss": 0.5671,
+      "step": 5222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2314839368385135,
+      "learning_rate": 1.3482171900948765e-05,
+      "loss": 0.5908,
+      "step": 5223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2438119786601036,
+      "learning_rate": 1.3479816378293662e-05,
+      "loss": 0.6246,
+      "step": 5224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0880687300732523,
+      "learning_rate": 1.3477460635939799e-05,
+      "loss": 0.5579,
+      "step": 5225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.138819323459468,
+      "learning_rate": 1.3475104674035902e-05,
+      "loss": 0.5878,
+      "step": 5226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2050894515681492,
+      "learning_rate": 1.347274849273072e-05,
+      "loss": 0.5829,
+      "step": 5227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2568860166304936,
+      "learning_rate": 1.3470392092173008e-05,
+      "loss": 0.6067,
+      "step": 5228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.251285178407395,
+      "learning_rate": 1.3468035472511539e-05,
+      "loss": 0.6112,
+      "step": 5229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.165477021261257,
+      "learning_rate": 1.3465678633895096e-05,
+      "loss": 0.5608,
+      "step": 5230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.229551427503285,
+      "learning_rate": 1.3463321576472482e-05,
+      "loss": 0.5796,
+      "step": 5231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1709672082044502,
+      "learning_rate": 1.346096430039251e-05,
+      "loss": 0.5179,
+      "step": 5232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1369282113279262,
+      "learning_rate": 1.3458606805804005e-05,
+      "loss": 0.5679,
+      "step": 5233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1793896128482513,
+      "learning_rate": 1.3456249092855805e-05,
+      "loss": 0.6115,
+      "step": 5234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2023482354212447,
+      "learning_rate": 1.345389116169677e-05,
+      "loss": 0.6598,
+      "step": 5235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2172925499769187,
+      "learning_rate": 1.3451533012475765e-05,
+      "loss": 0.5765,
+      "step": 5236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.149692757654883,
+      "learning_rate": 1.3449174645341675e-05,
+      "loss": 0.5298,
+      "step": 5237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1926994434452942,
+      "learning_rate": 1.3446816060443395e-05,
+      "loss": 0.5698,
+      "step": 5238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0959700261040861,
+      "learning_rate": 1.344445725792983e-05,
+      "loss": 0.5416,
+      "step": 5239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2075576868750588,
+      "learning_rate": 1.3442098237949913e-05,
+      "loss": 0.5785,
+      "step": 5240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.178225033912204,
+      "learning_rate": 1.3439739000652569e-05,
+      "loss": 0.5528,
+      "step": 5241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.169026842835405,
+      "learning_rate": 1.343737954618676e-05,
+      "loss": 0.6329,
+      "step": 5242
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1740442670473006,
+      "learning_rate": 1.3435019874701444e-05,
+      "loss": 0.5754,
+      "step": 5243
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1784416339875603,
+      "learning_rate": 1.3432659986345602e-05,
+      "loss": 0.5779,
+      "step": 5244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2582109187541184,
+      "learning_rate": 1.3430299881268223e-05,
+      "loss": 0.6197,
+      "step": 5245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0791516046442569,
+      "learning_rate": 1.3427939559618314e-05,
+      "loss": 0.5075,
+      "step": 5246
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.141248349776647,
+      "learning_rate": 1.3425579021544896e-05,
+      "loss": 0.5397,
+      "step": 5247
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1102728635464842,
+      "learning_rate": 1.3423218267197005e-05,
+      "loss": 0.5453,
+      "step": 5248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2214566521934216,
+      "learning_rate": 1.3420857296723674e-05,
+      "loss": 0.5963,
+      "step": 5249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2928796835472165,
+      "learning_rate": 1.341849611027398e-05,
+      "loss": 0.6264,
+      "step": 5250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1855863410627516,
+      "learning_rate": 1.3416134707996987e-05,
+      "loss": 0.5444,
+      "step": 5251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1259374421495798,
+      "learning_rate": 1.3413773090041786e-05,
+      "loss": 0.5365,
+      "step": 5252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2129149503618208,
+      "learning_rate": 1.3411411256557476e-05,
+      "loss": 0.5945,
+      "step": 5253
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1577300831319854,
+      "learning_rate": 1.3409049207693172e-05,
+      "loss": 0.5292,
+      "step": 5254
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.033072335342668,
+      "learning_rate": 1.3406686943598003e-05,
+      "loss": 0.527,
+      "step": 5255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2454299833050129,
+      "learning_rate": 1.3404324464421113e-05,
+      "loss": 0.6421,
+      "step": 5256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0841114416480866,
+      "learning_rate": 1.3401961770311655e-05,
+      "loss": 0.5374,
+      "step": 5257
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1010238196094357,
+      "learning_rate": 1.3399598861418798e-05,
+      "loss": 0.5386,
+      "step": 5258
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2353411888923391,
+      "learning_rate": 1.3397235737891726e-05,
+      "loss": 0.6201,
+      "step": 5259
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.173044816103312,
+      "learning_rate": 1.3394872399879634e-05,
+      "loss": 0.5306,
+      "step": 5260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3148521055884241,
+      "learning_rate": 1.339250884753173e-05,
+      "loss": 0.5534,
+      "step": 5261
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1990808165311708,
+      "learning_rate": 1.3390145080997243e-05,
+      "loss": 0.5092,
+      "step": 5262
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.111259096277895,
+      "learning_rate": 1.3387781100425407e-05,
+      "loss": 0.577,
+      "step": 5263
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1874351483756493,
+      "learning_rate": 1.3385416905965469e-05,
+      "loss": 0.5754,
+      "step": 5264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1739894356227907,
+      "learning_rate": 1.3383052497766695e-05,
+      "loss": 0.5689,
+      "step": 5265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.182610131391522,
+      "learning_rate": 1.3380687875978362e-05,
+      "loss": 0.5888,
+      "step": 5266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1114020297880585,
+      "learning_rate": 1.3378323040749764e-05,
+      "loss": 0.5546,
+      "step": 5267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3920953242723582,
+      "learning_rate": 1.3375957992230201e-05,
+      "loss": 0.604,
+      "step": 5268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.099141375347361,
+      "learning_rate": 1.3373592730568992e-05,
+      "loss": 0.5828,
+      "step": 5269
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1208625049883711,
+      "learning_rate": 1.337122725591547e-05,
+      "loss": 0.5746,
+      "step": 5270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.187791487405812,
+      "learning_rate": 1.3368861568418974e-05,
+      "loss": 0.6015,
+      "step": 5271
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2439445209305997,
+      "learning_rate": 1.3366495668228869e-05,
+      "loss": 0.5979,
+      "step": 5272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2590533468141456,
+      "learning_rate": 1.3364129555494524e-05,
+      "loss": 0.5534,
+      "step": 5273
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2511264017436263,
+      "learning_rate": 1.3361763230365323e-05,
+      "loss": 0.6046,
+      "step": 5274
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0830822188991507,
+      "learning_rate": 1.3359396692990667e-05,
+      "loss": 0.5522,
+      "step": 5275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2390720958747032,
+      "learning_rate": 1.3357029943519961e-05,
+      "loss": 0.6104,
+      "step": 5276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.150432592380982,
+      "learning_rate": 1.335466298210264e-05,
+      "loss": 0.5587,
+      "step": 5277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2277011809806118,
+      "learning_rate": 1.3352295808888134e-05,
+      "loss": 0.5624,
+      "step": 5278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1422140470621214,
+      "learning_rate": 1.3349928424025897e-05,
+      "loss": 0.5466,
+      "step": 5279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.09349253212638,
+      "learning_rate": 1.3347560827665398e-05,
+      "loss": 0.5071,
+      "step": 5280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.223498216001818,
+      "learning_rate": 1.3345193019956111e-05,
+      "loss": 0.6364,
+      "step": 5281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.289006966059338,
+      "learning_rate": 1.334282500104753e-05,
+      "loss": 0.5651,
+      "step": 5282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.18752660219612,
+      "learning_rate": 1.334045677108916e-05,
+      "loss": 0.6393,
+      "step": 5283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.158542525431692,
+      "learning_rate": 1.3338088330230518e-05,
+      "loss": 0.5726,
+      "step": 5284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.197256464713122,
+      "learning_rate": 1.3335719678621139e-05,
+      "loss": 0.5801,
+      "step": 5285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1925979907483184,
+      "learning_rate": 1.3333350816410568e-05,
+      "loss": 0.5056,
+      "step": 5286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1867715207308893,
+      "learning_rate": 1.3330981743748362e-05,
+      "loss": 0.5801,
+      "step": 5287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.229042995287446,
+      "learning_rate": 1.3328612460784091e-05,
+      "loss": 0.5919,
+      "step": 5288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2551172890891202,
+      "learning_rate": 1.3326242967667342e-05,
+      "loss": 0.5564,
+      "step": 5289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1319927011564592,
+      "learning_rate": 1.3323873264547713e-05,
+      "loss": 0.5559,
+      "step": 5290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.097894554470465,
+      "learning_rate": 1.3321503351574819e-05,
+      "loss": 0.5697,
+      "step": 5291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.251819954631812,
+      "learning_rate": 1.331913322889828e-05,
+      "loss": 0.5668,
+      "step": 5292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1211037452597765,
+      "learning_rate": 1.3316762896667737e-05,
+      "loss": 0.5977,
+      "step": 5293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1597795484971984,
+      "learning_rate": 1.3314392355032837e-05,
+      "loss": 0.6342,
+      "step": 5294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2404239059166946,
+      "learning_rate": 1.3312021604143247e-05,
+      "loss": 0.5782,
+      "step": 5295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1565144210754121,
+      "learning_rate": 1.3309650644148648e-05,
+      "loss": 0.5637,
+      "step": 5296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1266292324750615,
+      "learning_rate": 1.3307279475198729e-05,
+      "loss": 0.5848,
+      "step": 5297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2360664560822057,
+      "learning_rate": 1.3304908097443194e-05,
+      "loss": 0.5793,
+      "step": 5298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.162907738090381,
+      "learning_rate": 1.3302536511031755e-05,
+      "loss": 0.5645,
+      "step": 5299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1384226295574347,
+      "learning_rate": 1.3300164716114151e-05,
+      "loss": 0.5372,
+      "step": 5300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1184305065157014,
+      "learning_rate": 1.3297792712840121e-05,
+      "loss": 0.5599,
+      "step": 5301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0984647223842157,
+      "learning_rate": 1.3295420501359424e-05,
+      "loss": 0.5585,
+      "step": 5302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.303668966582131,
+      "learning_rate": 1.3293048081821827e-05,
+      "loss": 0.5829,
+      "step": 5303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3968011243431138,
+      "learning_rate": 1.3290675454377114e-05,
+      "loss": 0.6195,
+      "step": 5304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2358054069987254,
+      "learning_rate": 1.3288302619175082e-05,
+      "loss": 0.562,
+      "step": 5305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1645383950140562,
+      "learning_rate": 1.3285929576365541e-05,
+      "loss": 0.5729,
+      "step": 5306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.176074374847541,
+      "learning_rate": 1.3283556326098315e-05,
+      "loss": 0.5978,
+      "step": 5307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0618518647653676,
+      "learning_rate": 1.3281182868523235e-05,
+      "loss": 0.5393,
+      "step": 5308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1610846609746257,
+      "learning_rate": 1.327880920379015e-05,
+      "loss": 0.5998,
+      "step": 5309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.11523271817754,
+      "learning_rate": 1.3276435332048924e-05,
+      "loss": 0.5493,
+      "step": 5310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.141820672825066,
+      "learning_rate": 1.3274061253449434e-05,
+      "loss": 0.5584,
+      "step": 5311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1834210058875512,
+      "learning_rate": 1.3271686968141566e-05,
+      "loss": 0.5643,
+      "step": 5312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.14208984375,
+      "learning_rate": 1.3269312476275216e-05,
+      "loss": 0.5913,
+      "step": 5313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3027190017365033,
+      "learning_rate": 1.3266937778000304e-05,
+      "loss": 0.5957,
+      "step": 5314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.251638435412809,
+      "learning_rate": 1.3264562873466753e-05,
+      "loss": 0.5749,
+      "step": 5315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1748023029154155,
+      "learning_rate": 1.3262187762824507e-05,
+      "loss": 0.5329,
+      "step": 5316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.083255887319949,
+      "learning_rate": 1.3259812446223519e-05,
+      "loss": 0.529,
+      "step": 5317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2883288983658967,
+      "learning_rate": 1.325743692381375e-05,
+      "loss": 0.5766,
+      "step": 5318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.259754080312896,
+      "learning_rate": 1.3255061195745182e-05,
+      "loss": 0.5885,
+      "step": 5319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1492570820113677,
+      "learning_rate": 1.3252685262167803e-05,
+      "loss": 0.5734,
+      "step": 5320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2455881940779543,
+      "learning_rate": 1.3250309123231629e-05,
+      "loss": 0.576,
+      "step": 5321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1059794510006804,
+      "learning_rate": 1.3247932779086668e-05,
+      "loss": 0.57,
+      "step": 5322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2401644949863038,
+      "learning_rate": 1.324555622988295e-05,
+      "loss": 0.5872,
+      "step": 5323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.210066549819751,
+      "learning_rate": 1.3243179475770527e-05,
+      "loss": 0.6203,
+      "step": 5324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2145093844079853,
+      "learning_rate": 1.324080251689945e-05,
+      "loss": 0.5659,
+      "step": 5325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1001292434539458,
+      "learning_rate": 1.3238425353419788e-05,
+      "loss": 0.5789,
+      "step": 5326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1445031276541822,
+      "learning_rate": 1.3236047985481628e-05,
+      "loss": 0.5317,
+      "step": 5327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0828071869338616,
+      "learning_rate": 1.3233670413235059e-05,
+      "loss": 0.5317,
+      "step": 5328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2630660478292075,
+      "learning_rate": 1.3231292636830198e-05,
+      "loss": 0.5649,
+      "step": 5329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1830154869633025,
+      "learning_rate": 1.3228914656417156e-05,
+      "loss": 0.5365,
+      "step": 5330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2508666372133053,
+      "learning_rate": 1.3226536472146076e-05,
+      "loss": 0.5445,
+      "step": 5331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0856833400654309,
+      "learning_rate": 1.3224158084167104e-05,
+      "loss": 0.5069,
+      "step": 5332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.23172819078547,
+      "learning_rate": 1.3221779492630393e-05,
+      "loss": 0.632,
+      "step": 5333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2798880339617424,
+      "learning_rate": 1.3219400697686125e-05,
+      "loss": 0.5986,
+      "step": 5334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.200395584468608,
+      "learning_rate": 1.3217021699484476e-05,
+      "loss": 0.5474,
+      "step": 5335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1187671617438322,
+      "learning_rate": 1.3214642498175654e-05,
+      "loss": 0.5266,
+      "step": 5336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.244319693286147,
+      "learning_rate": 1.3212263093909865e-05,
+      "loss": 0.5722,
+      "step": 5337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2204440643076186,
+      "learning_rate": 1.320988348683733e-05,
+      "loss": 0.5923,
+      "step": 5338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.070028893980839,
+      "learning_rate": 1.3207503677108294e-05,
+      "loss": 0.5591,
+      "step": 5339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2247816144168766,
+      "learning_rate": 1.3205123664872999e-05,
+      "loss": 0.6196,
+      "step": 5340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2793022330368473,
+      "learning_rate": 1.3202743450281714e-05,
+      "loss": 0.5531,
+      "step": 5341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1453216219300684,
+      "learning_rate": 1.3200363033484709e-05,
+      "loss": 0.552,
+      "step": 5342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1078184786744194,
+      "learning_rate": 1.3197982414632272e-05,
+      "loss": 0.5344,
+      "step": 5343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2137208023252337,
+      "learning_rate": 1.319560159387471e-05,
+      "loss": 0.6294,
+      "step": 5344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0982733787029875,
+      "learning_rate": 1.3193220571362328e-05,
+      "loss": 0.561,
+      "step": 5345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1329781772872085,
+      "learning_rate": 1.319083934724546e-05,
+      "loss": 0.5656,
+      "step": 5346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2756605344661,
+      "learning_rate": 1.3188457921674443e-05,
+      "loss": 0.6146,
+      "step": 5347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3251142596516732,
+      "learning_rate": 1.3186076294799624e-05,
+      "loss": 0.6147,
+      "step": 5348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1498387928630491,
+      "learning_rate": 1.318369446677137e-05,
+      "loss": 0.5602,
+      "step": 5349
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2273020853198169,
+      "learning_rate": 1.3181312437740059e-05,
+      "loss": 0.5581,
+      "step": 5350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2477606742217344,
+      "learning_rate": 1.3178930207856082e-05,
+      "loss": 0.6612,
+      "step": 5351
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1590274752634329,
+      "learning_rate": 1.3176547777269838e-05,
+      "loss": 0.5668,
+      "step": 5352
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1709371756030678,
+      "learning_rate": 1.3174165146131746e-05,
+      "loss": 0.5459,
+      "step": 5353
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1294575965118923,
+      "learning_rate": 1.317178231459223e-05,
+      "loss": 0.5451,
+      "step": 5354
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1875674579935054,
+      "learning_rate": 1.3169399282801731e-05,
+      "loss": 0.5504,
+      "step": 5355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1710273219959833,
+      "learning_rate": 1.3167016050910709e-05,
+      "loss": 0.5638,
+      "step": 5356
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1267822770801894,
+      "learning_rate": 1.316463261906962e-05,
+      "loss": 0.5479,
+      "step": 5357
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0659092454120829,
+      "learning_rate": 1.3162248987428945e-05,
+      "loss": 0.5677,
+      "step": 5358
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.077452684943488,
+      "learning_rate": 1.3159865156139177e-05,
+      "loss": 0.5328,
+      "step": 5359
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1878566708238993,
+      "learning_rate": 1.3157481125350819e-05,
+      "loss": 0.5938,
+      "step": 5360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.136143699516451,
+      "learning_rate": 1.315509689521439e-05,
+      "loss": 0.5728,
+      "step": 5361
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2602971811117896,
+      "learning_rate": 1.3152712465880415e-05,
+      "loss": 0.5448,
+      "step": 5362
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1265945260711152,
+      "learning_rate": 1.3150327837499433e-05,
+      "loss": 0.5356,
+      "step": 5363
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1729767261710076,
+      "learning_rate": 1.3147943010222001e-05,
+      "loss": 0.6101,
+      "step": 5364
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1346307828696363,
+      "learning_rate": 1.3145557984198689e-05,
+      "loss": 0.5428,
+      "step": 5365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3409336650636199,
+      "learning_rate": 1.3143172759580072e-05,
+      "loss": 0.6448,
+      "step": 5366
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1121967052262058,
+      "learning_rate": 1.3140787336516743e-05,
+      "loss": 0.5277,
+      "step": 5367
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2291394516254261,
+      "learning_rate": 1.3138401715159301e-05,
+      "loss": 0.5759,
+      "step": 5368
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1726890788098938,
+      "learning_rate": 1.313601589565837e-05,
+      "loss": 0.6022,
+      "step": 5369
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1335142921452013,
+      "learning_rate": 1.3133629878164573e-05,
+      "loss": 0.5853,
+      "step": 5370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1698067278867612,
+      "learning_rate": 1.3131243662828558e-05,
+      "loss": 0.5806,
+      "step": 5371
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2313468585184584,
+      "learning_rate": 1.3128857249800975e-05,
+      "loss": 0.5683,
+      "step": 5372
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.251265409833926,
+      "learning_rate": 1.3126470639232487e-05,
+      "loss": 0.5452,
+      "step": 5373
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3091814685634704,
+      "learning_rate": 1.312408383127378e-05,
+      "loss": 0.6112,
+      "step": 5374
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.237139924963708,
+      "learning_rate": 1.3121696826075542e-05,
+      "loss": 0.5878,
+      "step": 5375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1172732340321048,
+      "learning_rate": 1.3119309623788479e-05,
+      "loss": 0.5318,
+      "step": 5376
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0744853295824257,
+      "learning_rate": 1.3116922224563306e-05,
+      "loss": 0.5582,
+      "step": 5377
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2414673451105471,
+      "learning_rate": 1.311453462855075e-05,
+      "loss": 0.5564,
+      "step": 5378
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.154450537043692,
+      "learning_rate": 1.3112146835901552e-05,
+      "loss": 0.6046,
+      "step": 5379
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0977557805777245,
+      "learning_rate": 1.3109758846766472e-05,
+      "loss": 0.5358,
+      "step": 5380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1126979437474374,
+      "learning_rate": 1.310737066129627e-05,
+      "loss": 0.555,
+      "step": 5381
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1137912050753107,
+      "learning_rate": 1.3104982279641727e-05,
+      "loss": 0.5925,
+      "step": 5382
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1289658580209436,
+      "learning_rate": 1.3102593701953632e-05,
+      "loss": 0.5432,
+      "step": 5383
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2640440686064511,
+      "learning_rate": 1.3100204928382787e-05,
+      "loss": 0.6193,
+      "step": 5384
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1640846711165926,
+      "learning_rate": 1.3097815959080013e-05,
+      "loss": 0.5981,
+      "step": 5385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2061643481110838,
+      "learning_rate": 1.3095426794196136e-05,
+      "loss": 0.5771,
+      "step": 5386
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1018096430150346,
+      "learning_rate": 1.3093037433881995e-05,
+      "loss": 0.5421,
+      "step": 5387
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1087404504857186,
+      "learning_rate": 1.3090647878288441e-05,
+      "loss": 0.5428,
+      "step": 5388
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.328570481928475,
+      "learning_rate": 1.3088258127566338e-05,
+      "loss": 0.6409,
+      "step": 5389
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2403104503457993,
+      "learning_rate": 1.3085868181866571e-05,
+      "loss": 0.598,
+      "step": 5390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2386832081238086,
+      "learning_rate": 1.3083478041340023e-05,
+      "loss": 0.5799,
+      "step": 5391
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1597264068335127,
+      "learning_rate": 1.3081087706137596e-05,
+      "loss": 0.608,
+      "step": 5392
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1772125758332481,
+      "learning_rate": 1.307869717641021e-05,
+      "loss": 0.559,
+      "step": 5393
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0811032057766397,
+      "learning_rate": 1.3076306452308782e-05,
+      "loss": 0.5869,
+      "step": 5394
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.205162403986662,
+      "learning_rate": 1.3073915533984262e-05,
+      "loss": 0.5905,
+      "step": 5395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2341183262763087,
+      "learning_rate": 1.307152442158759e-05,
+      "loss": 0.613,
+      "step": 5396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1664420263548732,
+      "learning_rate": 1.3069133115269734e-05,
+      "loss": 0.5878,
+      "step": 5397
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0628185355719757,
+      "learning_rate": 1.3066741615181675e-05,
+      "loss": 0.5476,
+      "step": 5398
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.218004585605126,
+      "learning_rate": 1.306434992147439e-05,
+      "loss": 0.6091,
+      "step": 5399
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.090991984274113,
+      "learning_rate": 1.306195803429889e-05,
+      "loss": 0.4964,
+      "step": 5400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.113897588008314,
+      "learning_rate": 1.3059565953806177e-05,
+      "loss": 0.5366,
+      "step": 5401
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2068524274488157,
+      "learning_rate": 1.305717368014728e-05,
+      "loss": 0.6072,
+      "step": 5402
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.210839987489532,
+      "learning_rate": 1.3054781213473238e-05,
+      "loss": 0.5776,
+      "step": 5403
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1703435427695652,
+      "learning_rate": 1.3052388553935096e-05,
+      "loss": 0.5825,
+      "step": 5404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.168563844267776,
+      "learning_rate": 1.304999570168392e-05,
+      "loss": 0.5624,
+      "step": 5405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2287186544306676,
+      "learning_rate": 1.3047602656870775e-05,
+      "loss": 0.5939,
+      "step": 5406
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2171640592186117,
+      "learning_rate": 1.3045209419646749e-05,
+      "loss": 0.594,
+      "step": 5407
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1286022476537598,
+      "learning_rate": 1.3042815990162944e-05,
+      "loss": 0.573,
+      "step": 5408
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1666646968734098,
+      "learning_rate": 1.3040422368570466e-05,
+      "loss": 0.5778,
+      "step": 5409
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1808447813898302,
+      "learning_rate": 1.3038028555020444e-05,
+      "loss": 0.5682,
+      "step": 5410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1902489716674234,
+      "learning_rate": 1.3035634549663995e-05,
+      "loss": 0.5291,
+      "step": 5411
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1010248481863851,
+      "learning_rate": 1.3033240352652281e-05,
+      "loss": 0.5467,
+      "step": 5412
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.274650557633351,
+      "learning_rate": 1.3030845964136452e-05,
+      "loss": 0.6078,
+      "step": 5413
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1680105960966254,
+      "learning_rate": 1.3028451384267679e-05,
+      "loss": 0.5619,
+      "step": 5414
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1307461011483286,
+      "learning_rate": 1.3026056613197151e-05,
+      "loss": 0.5746,
+      "step": 5415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1679891118855061,
+      "learning_rate": 1.3023661651076051e-05,
+      "loss": 0.5522,
+      "step": 5416
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2030811549532923,
+      "learning_rate": 1.3021266498055592e-05,
+      "loss": 0.5574,
+      "step": 5417
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2734235517053876,
+      "learning_rate": 1.3018871154286991e-05,
+      "loss": 0.6376,
+      "step": 5418
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1515076270933733,
+      "learning_rate": 1.3016475619921477e-05,
+      "loss": 0.5837,
+      "step": 5419
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2972100985439499,
+      "learning_rate": 1.3014079895110299e-05,
+      "loss": 0.6417,
+      "step": 5420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1155738647880546,
+      "learning_rate": 1.3011683980004705e-05,
+      "loss": 0.5376,
+      "step": 5421
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1662838625719312,
+      "learning_rate": 1.3009287874755963e-05,
+      "loss": 0.6136,
+      "step": 5422
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1438499740515875,
+      "learning_rate": 1.3006891579515351e-05,
+      "loss": 0.528,
+      "step": 5423
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1405953181337083,
+      "learning_rate": 1.3004495094434157e-05,
+      "loss": 0.571,
+      "step": 5424
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1564098453865295,
+      "learning_rate": 1.3002098419663692e-05,
+      "loss": 0.5731,
+      "step": 5425
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0746658781804426,
+      "learning_rate": 1.2999701555355264e-05,
+      "loss": 0.5333,
+      "step": 5426
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1580312257157257,
+      "learning_rate": 1.2997304501660197e-05,
+      "loss": 0.5383,
+      "step": 5427
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1553913227213064,
+      "learning_rate": 1.2994907258729835e-05,
+      "loss": 0.5779,
+      "step": 5428
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1828487053070678,
+      "learning_rate": 1.2992509826715525e-05,
+      "loss": 0.5551,
+      "step": 5429
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.202424030957592,
+      "learning_rate": 1.2990112205768632e-05,
+      "loss": 0.6345,
+      "step": 5430
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2214735849829235,
+      "learning_rate": 1.2987714396040527e-05,
+      "loss": 0.6044,
+      "step": 5431
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1452872738005164,
+      "learning_rate": 1.2985316397682597e-05,
+      "loss": 0.5728,
+      "step": 5432
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.097056486325496,
+      "learning_rate": 1.2982918210846243e-05,
+      "loss": 0.542,
+      "step": 5433
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.077531125798468,
+      "learning_rate": 1.298051983568287e-05,
+      "loss": 0.5364,
+      "step": 5434
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2556836611923696,
+      "learning_rate": 1.2978121272343904e-05,
+      "loss": 0.5945,
+      "step": 5435
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.165231003022694,
+      "learning_rate": 1.2975722520980777e-05,
+      "loss": 0.5832,
+      "step": 5436
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2382334510410542,
+      "learning_rate": 1.2973323581744935e-05,
+      "loss": 0.5825,
+      "step": 5437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2646124286017335,
+      "learning_rate": 1.2970924454787834e-05,
+      "loss": 0.5928,
+      "step": 5438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.168861634741927,
+      "learning_rate": 1.2968525140260946e-05,
+      "loss": 0.6037,
+      "step": 5439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.072208034726781,
+      "learning_rate": 1.2966125638315745e-05,
+      "loss": 0.5404,
+      "step": 5440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0987807517581774,
+      "learning_rate": 1.2963725949103734e-05,
+      "loss": 0.5726,
+      "step": 5441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1791842062806022,
+      "learning_rate": 1.2961326072776412e-05,
+      "loss": 0.5786,
+      "step": 5442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2925609394958855,
+      "learning_rate": 1.2958926009485297e-05,
+      "loss": 0.5977,
+      "step": 5443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2435406686032486,
+      "learning_rate": 1.2956525759381917e-05,
+      "loss": 0.6286,
+      "step": 5444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1867145148999685,
+      "learning_rate": 1.295412532261781e-05,
+      "loss": 0.542,
+      "step": 5445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0957595348287095,
+      "learning_rate": 1.2951724699344532e-05,
+      "loss": 0.5126,
+      "step": 5446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2980519664571566,
+      "learning_rate": 1.2949323889713643e-05,
+      "loss": 0.6512,
+      "step": 5447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2518532366486848,
+      "learning_rate": 1.2946922893876722e-05,
+      "loss": 0.5531,
+      "step": 5448
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0681618572413925,
+      "learning_rate": 1.2944521711985357e-05,
+      "loss": 0.5069,
+      "step": 5449
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2524750048794988,
+      "learning_rate": 1.2942120344191141e-05,
+      "loss": 0.5952,
+      "step": 5450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1794976087920306,
+      "learning_rate": 1.293971879064569e-05,
+      "loss": 0.5681,
+      "step": 5451
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0510496909846871,
+      "learning_rate": 1.2937317051500622e-05,
+      "loss": 0.5169,
+      "step": 5452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2114186469301964,
+      "learning_rate": 1.2934915126907575e-05,
+      "loss": 0.5952,
+      "step": 5453
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1396633431604761,
+      "learning_rate": 1.2932513017018197e-05,
+      "loss": 0.5781,
+      "step": 5454
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1030882421318362,
+      "learning_rate": 1.293011072198414e-05,
+      "loss": 0.5656,
+      "step": 5455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1908544544583044,
+      "learning_rate": 1.2927708241957077e-05,
+      "loss": 0.5282,
+      "step": 5456
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1783158873669861,
+      "learning_rate": 1.2925305577088687e-05,
+      "loss": 0.5524,
+      "step": 5457
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1286463982909718,
+      "learning_rate": 1.2922902727530663e-05,
+      "loss": 0.5551,
+      "step": 5458
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1902179733162859,
+      "learning_rate": 1.2920499693434712e-05,
+      "loss": 0.6409,
+      "step": 5459
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2854356548982162,
+      "learning_rate": 1.2918096474952544e-05,
+      "loss": 0.6034,
+      "step": 5460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.156583892451261,
+      "learning_rate": 1.2915693072235893e-05,
+      "loss": 0.5793,
+      "step": 5461
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2416669126591449,
+      "learning_rate": 1.2913289485436492e-05,
+      "loss": 0.5335,
+      "step": 5462
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1119004637883199,
+      "learning_rate": 1.2910885714706093e-05,
+      "loss": 0.5481,
+      "step": 5463
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2679661419948003,
+      "learning_rate": 1.290848176019647e-05,
+      "loss": 0.6347,
+      "step": 5464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2130405990586628,
+      "learning_rate": 1.290607762205938e-05,
+      "loss": 0.6087,
+      "step": 5465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1688536286948465,
+      "learning_rate": 1.2903673300446623e-05,
+      "loss": 0.5916,
+      "step": 5466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.303191646637205,
+      "learning_rate": 1.2901268795509982e-05,
+      "loss": 0.6474,
+      "step": 5467
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.100066284870186,
+      "learning_rate": 1.2898864107401275e-05,
+      "loss": 0.5632,
+      "step": 5468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2421815500176878,
+      "learning_rate": 1.2896459236272325e-05,
+      "loss": 0.5645,
+      "step": 5469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1206938133238493,
+      "learning_rate": 1.2894054182274956e-05,
+      "loss": 0.5103,
+      "step": 5470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1077585398199128,
+      "learning_rate": 1.2891648945561017e-05,
+      "loss": 0.5008,
+      "step": 5471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.185392969941593,
+      "learning_rate": 1.2889243526282357e-05,
+      "loss": 0.564,
+      "step": 5472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1988177296873601,
+      "learning_rate": 1.288683792459085e-05,
+      "loss": 0.5892,
+      "step": 5473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2340225972157672,
+      "learning_rate": 1.2884432140638375e-05,
+      "loss": 0.6147,
+      "step": 5474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2384168869486223,
+      "learning_rate": 1.2882026174576812e-05,
+      "loss": 0.5768,
+      "step": 5475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1702327665458885,
+      "learning_rate": 1.2879620026558067e-05,
+      "loss": 0.5321,
+      "step": 5476
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0680964005035958,
+      "learning_rate": 1.2877213696734052e-05,
+      "loss": 0.5685,
+      "step": 5477
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.133198376408608,
+      "learning_rate": 1.2874807185256692e-05,
+      "loss": 0.5319,
+      "step": 5478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2608793791241737,
+      "learning_rate": 1.2872400492277928e-05,
+      "loss": 0.5705,
+      "step": 5479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3037042624770598,
+      "learning_rate": 1.2869993617949696e-05,
+      "loss": 0.6176,
+      "step": 5480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2906044725166501,
+      "learning_rate": 1.2867586562423962e-05,
+      "loss": 0.5741,
+      "step": 5481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1569897115541745,
+      "learning_rate": 1.2865179325852693e-05,
+      "loss": 0.5376,
+      "step": 5482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2820102018108113,
+      "learning_rate": 1.2862771908387867e-05,
+      "loss": 0.6216,
+      "step": 5483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2591156931585912,
+      "learning_rate": 1.2860364310181488e-05,
+      "loss": 0.5675,
+      "step": 5484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1636384537255418,
+      "learning_rate": 1.2857956531385548e-05,
+      "loss": 0.579,
+      "step": 5485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1578285679493767,
+      "learning_rate": 1.2855548572152066e-05,
+      "loss": 0.5645,
+      "step": 5486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1476531172177928,
+      "learning_rate": 1.2853140432633074e-05,
+      "loss": 0.503,
+      "step": 5487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2773189717260567,
+      "learning_rate": 1.2850732112980602e-05,
+      "loss": 0.5681,
+      "step": 5488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2145730357881368,
+      "learning_rate": 1.2848323613346708e-05,
+      "loss": 0.5966,
+      "step": 5489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1647737109736245,
+      "learning_rate": 1.2845914933883443e-05,
+      "loss": 0.6117,
+      "step": 5490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1003217811702242,
+      "learning_rate": 1.2843506074742888e-05,
+      "loss": 0.5412,
+      "step": 5491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.202108870598096,
+      "learning_rate": 1.2841097036077125e-05,
+      "loss": 0.615,
+      "step": 5492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1819249339173739,
+      "learning_rate": 1.283868781803825e-05,
+      "loss": 0.5932,
+      "step": 5493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0883979641669264,
+      "learning_rate": 1.2836278420778366e-05,
+      "loss": 0.5207,
+      "step": 5494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1347918351473658,
+      "learning_rate": 1.2833868844449588e-05,
+      "loss": 0.5725,
+      "step": 5495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.232446201767509,
+      "learning_rate": 1.283145908920405e-05,
+      "loss": 0.592,
+      "step": 5496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3793116937930812,
+      "learning_rate": 1.2829049155193896e-05,
+      "loss": 0.6154,
+      "step": 5497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.246839533327724,
+      "learning_rate": 1.282663904257127e-05,
+      "loss": 0.6238,
+      "step": 5498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0381274497630815,
+      "learning_rate": 1.2824228751488339e-05,
+      "loss": 0.505,
+      "step": 5499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.212783194435812,
+      "learning_rate": 1.2821818282097273e-05,
+      "loss": 0.5273,
+      "step": 5500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1216896945427937,
+      "learning_rate": 1.281940763455026e-05,
+      "loss": 0.5618,
+      "step": 5501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3120100378402657,
+      "learning_rate": 1.28169968089995e-05,
+      "loss": 0.5818,
+      "step": 5502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1351507849670044,
+      "learning_rate": 1.2814585805597197e-05,
+      "loss": 0.5136,
+      "step": 5503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2382589633031131,
+      "learning_rate": 1.2812174624495569e-05,
+      "loss": 0.5979,
+      "step": 5504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1588280264088582,
+      "learning_rate": 1.2809763265846851e-05,
+      "loss": 0.5037,
+      "step": 5505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1998168249197858,
+      "learning_rate": 1.2807351729803277e-05,
+      "loss": 0.5475,
+      "step": 5506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2553727555384517,
+      "learning_rate": 1.280494001651711e-05,
+      "loss": 0.6203,
+      "step": 5507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2009800703537514,
+      "learning_rate": 1.2802528126140604e-05,
+      "loss": 0.6095,
+      "step": 5508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2123136514830408,
+      "learning_rate": 1.2800116058826037e-05,
+      "loss": 0.6154,
+      "step": 5509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1539214632326678,
+      "learning_rate": 1.2797703814725702e-05,
+      "loss": 0.576,
+      "step": 5510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1659747297057943,
+      "learning_rate": 1.2795291393991885e-05,
+      "loss": 0.6005,
+      "step": 5511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1356948488428176,
+      "learning_rate": 1.2792878796776904e-05,
+      "loss": 0.6047,
+      "step": 5512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.085219605630756,
+      "learning_rate": 1.2790466023233075e-05,
+      "loss": 0.5306,
+      "step": 5513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2194775586036035,
+      "learning_rate": 1.2788053073512728e-05,
+      "loss": 0.6436,
+      "step": 5514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1882131342579478,
+      "learning_rate": 1.2785639947768207e-05,
+      "loss": 0.5168,
+      "step": 5515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.133888943369136,
+      "learning_rate": 1.2783226646151863e-05,
+      "loss": 0.5642,
+      "step": 5516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.153933860123063,
+      "learning_rate": 1.2780813168816062e-05,
+      "loss": 0.5221,
+      "step": 5517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1556269023543317,
+      "learning_rate": 1.2778399515913177e-05,
+      "loss": 0.5666,
+      "step": 5518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.175304886646964,
+      "learning_rate": 1.2775985687595598e-05,
+      "loss": 0.5668,
+      "step": 5519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1536244145709227,
+      "learning_rate": 1.277357168401572e-05,
+      "loss": 0.4967,
+      "step": 5520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1527478560173565,
+      "learning_rate": 1.2771157505325951e-05,
+      "loss": 0.5899,
+      "step": 5521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.133982560553497,
+      "learning_rate": 1.2768743151678711e-05,
+      "loss": 0.5354,
+      "step": 5522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.19549359713714,
+      "learning_rate": 1.2766328623226434e-05,
+      "loss": 0.59,
+      "step": 5523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.160543820156621,
+      "learning_rate": 1.2763913920121554e-05,
+      "loss": 0.5791,
+      "step": 5524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2247717352747132,
+      "learning_rate": 1.2761499042516531e-05,
+      "loss": 0.5581,
+      "step": 5525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0866881595916071,
+      "learning_rate": 1.2759083990563825e-05,
+      "loss": 0.4941,
+      "step": 5526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1619143538764027,
+      "learning_rate": 1.2756668764415913e-05,
+      "loss": 0.5733,
+      "step": 5527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0770966974802965,
+      "learning_rate": 1.275425336422528e-05,
+      "loss": 0.5467,
+      "step": 5528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1545138340745242,
+      "learning_rate": 1.2751837790144419e-05,
+      "loss": 0.6053,
+      "step": 5529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1765427768872843,
+      "learning_rate": 1.2749422042325846e-05,
+      "loss": 0.5294,
+      "step": 5530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.301201921320522,
+      "learning_rate": 1.2747006120922068e-05,
+      "loss": 0.6116,
+      "step": 5531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1333565819450089,
+      "learning_rate": 1.2744590026085622e-05,
+      "loss": 0.5351,
+      "step": 5532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2013959970184338,
+      "learning_rate": 1.2742173757969052e-05,
+      "loss": 0.5668,
+      "step": 5533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1777583130891491,
+      "learning_rate": 1.2739757316724901e-05,
+      "loss": 0.521,
+      "step": 5534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.215725790125941,
+      "learning_rate": 1.2737340702505737e-05,
+      "loss": 0.5632,
+      "step": 5535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.22141585638205,
+      "learning_rate": 1.273492391546413e-05,
+      "loss": 0.5806,
+      "step": 5536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1856712012169837,
+      "learning_rate": 1.2732506955752665e-05,
+      "loss": 0.5805,
+      "step": 5537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1348465646054706,
+      "learning_rate": 1.2730089823523943e-05,
+      "loss": 0.564,
+      "step": 5538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3781617832248785,
+      "learning_rate": 1.2727672518930561e-05,
+      "loss": 0.6869,
+      "step": 5539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1141266942087853,
+      "learning_rate": 1.2725255042125142e-05,
+      "loss": 0.5721,
+      "step": 5540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1295708412728245,
+      "learning_rate": 1.2722837393260308e-05,
+      "loss": 0.5473,
+      "step": 5541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.19208789774343,
+      "learning_rate": 1.2720419572488705e-05,
+      "loss": 0.5541,
+      "step": 5542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2233611198737826,
+      "learning_rate": 1.2718001579962978e-05,
+      "loss": 0.5988,
+      "step": 5543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2402776275061307,
+      "learning_rate": 1.2715583415835788e-05,
+      "loss": 0.6028,
+      "step": 5544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0926991455351278,
+      "learning_rate": 1.2713165080259805e-05,
+      "loss": 0.5535,
+      "step": 5545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1117348622208707,
+      "learning_rate": 1.2710746573387716e-05,
+      "loss": 0.5454,
+      "step": 5546
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1983621128448163,
+      "learning_rate": 1.2708327895372208e-05,
+      "loss": 0.5822,
+      "step": 5547
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0697513975137425,
+      "learning_rate": 1.2705909046365987e-05,
+      "loss": 0.5298,
+      "step": 5548
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1547100534675536,
+      "learning_rate": 1.2703490026521766e-05,
+      "loss": 0.521,
+      "step": 5549
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1326668481255966,
+      "learning_rate": 1.2701070835992273e-05,
+      "loss": 0.5192,
+      "step": 5550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0812013383620682,
+      "learning_rate": 1.2698651474930239e-05,
+      "loss": 0.567,
+      "step": 5551
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1635503986592208,
+      "learning_rate": 1.269623194348842e-05,
+      "loss": 0.586,
+      "step": 5552
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1103739471997414,
+      "learning_rate": 1.2693812241819565e-05,
+      "loss": 0.508,
+      "step": 5553
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3577002370274034,
+      "learning_rate": 1.2691392370076443e-05,
+      "loss": 0.6356,
+      "step": 5554
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.099004982535233,
+      "learning_rate": 1.2688972328411836e-05,
+      "loss": 0.5487,
+      "step": 5555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.31773268886971,
+      "learning_rate": 1.2686552116978535e-05,
+      "loss": 0.6177,
+      "step": 5556
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1867376691403835,
+      "learning_rate": 1.2684131735929337e-05,
+      "loss": 0.5549,
+      "step": 5557
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1766972316486015,
+      "learning_rate": 1.2681711185417053e-05,
+      "loss": 0.5532,
+      "step": 5558
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1879420210693632,
+      "learning_rate": 1.2679290465594507e-05,
+      "loss": 0.5639,
+      "step": 5559
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1284058197569167,
+      "learning_rate": 1.2676869576614527e-05,
+      "loss": 0.5732,
+      "step": 5560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1950623992366058,
+      "learning_rate": 1.2674448518629964e-05,
+      "loss": 0.5867,
+      "step": 5561
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2503843193526816,
+      "learning_rate": 1.2672027291793669e-05,
+      "loss": 0.6329,
+      "step": 5562
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1781078651052352,
+      "learning_rate": 1.2669605896258503e-05,
+      "loss": 0.5366,
+      "step": 5563
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1897859655764778,
+      "learning_rate": 1.2667184332177342e-05,
+      "loss": 0.5648,
+      "step": 5564
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.254836690923467,
+      "learning_rate": 1.2664762599703073e-05,
+      "loss": 0.6093,
+      "step": 5565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2439129439922414,
+      "learning_rate": 1.2662340698988595e-05,
+      "loss": 0.5668,
+      "step": 5566
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.142804037077906,
+      "learning_rate": 1.2659918630186814e-05,
+      "loss": 0.5681,
+      "step": 5567
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1572250173051393,
+      "learning_rate": 1.2657496393450646e-05,
+      "loss": 0.5647,
+      "step": 5568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1247183129196163,
+      "learning_rate": 1.2655073988933016e-05,
+      "loss": 0.5784,
+      "step": 5569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1487916672780292,
+      "learning_rate": 1.2652651416786867e-05,
+      "loss": 0.5301,
+      "step": 5570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2677035277042292,
+      "learning_rate": 1.2650228677165153e-05,
+      "loss": 0.593,
+      "step": 5571
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.145158302991299,
+      "learning_rate": 1.2647805770220826e-05,
+      "loss": 0.5562,
+      "step": 5572
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1512460944226481,
+      "learning_rate": 1.264538269610686e-05,
+      "loss": 0.5712,
+      "step": 5573
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1257718934961527,
+      "learning_rate": 1.2642959454976236e-05,
+      "loss": 0.5677,
+      "step": 5574
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2000249581920426,
+      "learning_rate": 1.2640536046981943e-05,
+      "loss": 0.5582,
+      "step": 5575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2453488123253045,
+      "learning_rate": 1.263811247227699e-05,
+      "loss": 0.5999,
+      "step": 5576
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2416045542463858,
+      "learning_rate": 1.2635688731014386e-05,
+      "loss": 0.5865,
+      "step": 5577
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.110429880136358,
+      "learning_rate": 1.2633264823347156e-05,
+      "loss": 0.5823,
+      "step": 5578
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2268713786945737,
+      "learning_rate": 1.2630840749428327e-05,
+      "loss": 0.5936,
+      "step": 5579
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1897675798547742,
+      "learning_rate": 1.2628416509410947e-05,
+      "loss": 0.6204,
+      "step": 5580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1945103465734335,
+      "learning_rate": 1.2625992103448077e-05,
+      "loss": 0.584,
+      "step": 5581
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.129620547177039,
+      "learning_rate": 1.2623567531692774e-05,
+      "loss": 0.5422,
+      "step": 5582
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.106649139296138,
+      "learning_rate": 1.2621142794298118e-05,
+      "loss": 0.5495,
+      "step": 5583
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2286843091571198,
+      "learning_rate": 1.2618717891417194e-05,
+      "loss": 0.5841,
+      "step": 5584
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2267494302192081,
+      "learning_rate": 1.2616292823203098e-05,
+      "loss": 0.5958,
+      "step": 5585
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1938057172464853,
+      "learning_rate": 1.2613867589808939e-05,
+      "loss": 0.5393,
+      "step": 5586
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.253108260882443,
+      "learning_rate": 1.2611442191387836e-05,
+      "loss": 0.5691,
+      "step": 5587
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2112076488763577,
+      "learning_rate": 1.2609016628092907e-05,
+      "loss": 0.5426,
+      "step": 5588
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1524016802176051,
+      "learning_rate": 1.2606590900077303e-05,
+      "loss": 0.5564,
+      "step": 5589
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1014883679274898,
+      "learning_rate": 1.2604165007494164e-05,
+      "loss": 0.5093,
+      "step": 5590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9059208445048783,
+      "learning_rate": 1.2601738950496654e-05,
+      "loss": 0.5726,
+      "step": 5591
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0752109453691523,
+      "learning_rate": 1.2599312729237943e-05,
+      "loss": 0.5657,
+      "step": 5592
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1366389925003257,
+      "learning_rate": 1.2596886343871204e-05,
+      "loss": 0.5124,
+      "step": 5593
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1415359830661342,
+      "learning_rate": 1.2594459794549636e-05,
+      "loss": 0.5544,
+      "step": 5594
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2531306640040374,
+      "learning_rate": 1.2592033081426434e-05,
+      "loss": 0.5962,
+      "step": 5595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0777766452752768,
+      "learning_rate": 1.2589606204654809e-05,
+      "loss": 0.5125,
+      "step": 5596
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1206248828145364,
+      "learning_rate": 1.2587179164387987e-05,
+      "loss": 0.5679,
+      "step": 5597
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.157364849085201,
+      "learning_rate": 1.2584751960779192e-05,
+      "loss": 0.5925,
+      "step": 5598
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1152819407886694,
+      "learning_rate": 1.2582324593981673e-05,
+      "loss": 0.5686,
+      "step": 5599
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.298006047109851,
+      "learning_rate": 1.2579897064148678e-05,
+      "loss": 0.5408,
+      "step": 5600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.162330619203509,
+      "learning_rate": 1.2577469371433473e-05,
+      "loss": 0.5657,
+      "step": 5601
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1326334318498235,
+      "learning_rate": 1.2575041515989328e-05,
+      "loss": 0.5385,
+      "step": 5602
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2027095721500707,
+      "learning_rate": 1.2572613497969524e-05,
+      "loss": 0.5648,
+      "step": 5603
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2390831598048317,
+      "learning_rate": 1.257018531752736e-05,
+      "loss": 0.5448,
+      "step": 5604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2188423806415503,
+      "learning_rate": 1.2567756974816134e-05,
+      "loss": 0.5245,
+      "step": 5605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0975876104539766,
+      "learning_rate": 1.2565328469989165e-05,
+      "loss": 0.5449,
+      "step": 5606
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1708930924574363,
+      "learning_rate": 1.2562899803199773e-05,
+      "loss": 0.626,
+      "step": 5607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2815041638863705,
+      "learning_rate": 1.2560470974601294e-05,
+      "loss": 0.6245,
+      "step": 5608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2052114156399512,
+      "learning_rate": 1.255804198434707e-05,
+      "loss": 0.5591,
+      "step": 5609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1299411029995607,
+      "learning_rate": 1.2555612832590458e-05,
+      "loss": 0.5619,
+      "step": 5610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3317409881521767,
+      "learning_rate": 1.2553183519484826e-05,
+      "loss": 0.6291,
+      "step": 5611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0300141645587158,
+      "learning_rate": 1.2550754045183544e-05,
+      "loss": 0.5636,
+      "step": 5612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0895233230160215,
+      "learning_rate": 1.2548324409839999e-05,
+      "loss": 0.5437,
+      "step": 5613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2064299829641423,
+      "learning_rate": 1.2545894613607585e-05,
+      "loss": 0.6266,
+      "step": 5614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2018781864879817,
+      "learning_rate": 1.254346465663971e-05,
+      "loss": 0.6192,
+      "step": 5615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.288902272756921,
+      "learning_rate": 1.254103453908979e-05,
+      "loss": 0.6152,
+      "step": 5616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.189199335994593,
+      "learning_rate": 1.2538604261111247e-05,
+      "loss": 0.6015,
+      "step": 5617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0927091823364878,
+      "learning_rate": 1.253617382285752e-05,
+      "loss": 0.5194,
+      "step": 5618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1990401541959328,
+      "learning_rate": 1.2533743224482055e-05,
+      "loss": 0.5738,
+      "step": 5619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2292325546364247,
+      "learning_rate": 1.253131246613831e-05,
+      "loss": 0.5667,
+      "step": 5620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1838606731718777,
+      "learning_rate": 1.2528881547979748e-05,
+      "loss": 0.5645,
+      "step": 5621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.133026471154515,
+      "learning_rate": 1.2526450470159845e-05,
+      "loss": 0.5647,
+      "step": 5622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0842670244002264,
+      "learning_rate": 1.2524019232832089e-05,
+      "loss": 0.5356,
+      "step": 5623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1829979533212456,
+      "learning_rate": 1.252158783614998e-05,
+      "loss": 0.6011,
+      "step": 5624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2348087852342216,
+      "learning_rate": 1.2519156280267017e-05,
+      "loss": 0.602,
+      "step": 5625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1313194169301475,
+      "learning_rate": 1.2516724565336724e-05,
+      "loss": 0.5417,
+      "step": 5626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2771990399531177,
+      "learning_rate": 1.2514292691512624e-05,
+      "loss": 0.634,
+      "step": 5627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2768444053506123,
+      "learning_rate": 1.2511860658948252e-05,
+      "loss": 0.5424,
+      "step": 5628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0379950985140463,
+      "learning_rate": 1.250942846779716e-05,
+      "loss": 0.5099,
+      "step": 5629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1584306721760087,
+      "learning_rate": 1.2506996118212897e-05,
+      "loss": 0.5554,
+      "step": 5630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2251718011391377,
+      "learning_rate": 1.250456361034904e-05,
+      "loss": 0.6173,
+      "step": 5631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2020889379061062,
+      "learning_rate": 1.2502130944359161e-05,
+      "loss": 0.5743,
+      "step": 5632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.221823460499267,
+      "learning_rate": 1.249969812039684e-05,
+      "loss": 0.5637,
+      "step": 5633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.079684139086382,
+      "learning_rate": 1.2497265138615686e-05,
+      "loss": 0.538,
+      "step": 5634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2561091386763716,
+      "learning_rate": 1.2494831999169296e-05,
+      "loss": 0.5183,
+      "step": 5635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2243682283191835,
+      "learning_rate": 1.2492398702211293e-05,
+      "loss": 0.5708,
+      "step": 5636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1549289476742823,
+      "learning_rate": 1.2489965247895302e-05,
+      "loss": 0.5946,
+      "step": 5637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0833222987762312,
+      "learning_rate": 1.2487531636374954e-05,
+      "loss": 0.5671,
+      "step": 5638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.261160712356664,
+      "learning_rate": 1.24850978678039e-05,
+      "loss": 0.5592,
+      "step": 5639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2041147792861158,
+      "learning_rate": 1.2482663942335798e-05,
+      "loss": 0.5817,
+      "step": 5640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0903270019196591,
+      "learning_rate": 1.2480229860124313e-05,
+      "loss": 0.5514,
+      "step": 5641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2564350428399476,
+      "learning_rate": 1.2477795621323121e-05,
+      "loss": 0.5771,
+      "step": 5642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.164034849516926,
+      "learning_rate": 1.2475361226085907e-05,
+      "loss": 0.5226,
+      "step": 5643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2202677934690032,
+      "learning_rate": 1.2472926674566366e-05,
+      "loss": 0.5313,
+      "step": 5644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1276619465438138,
+      "learning_rate": 1.2470491966918205e-05,
+      "loss": 0.5251,
+      "step": 5645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.206963299563885,
+      "learning_rate": 1.2468057103295144e-05,
+      "loss": 0.5663,
+      "step": 5646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1818964405424834,
+      "learning_rate": 1.2465622083850903e-05,
+      "loss": 0.5999,
+      "step": 5647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3265614450043886,
+      "learning_rate": 1.2463186908739217e-05,
+      "loss": 0.593,
+      "step": 5648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1676976836663495,
+      "learning_rate": 1.2460751578113832e-05,
+      "loss": 0.5849,
+      "step": 5649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1147956586488383,
+      "learning_rate": 1.2458316092128509e-05,
+      "loss": 0.5933,
+      "step": 5650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1741671715205786,
+      "learning_rate": 1.2455880450937006e-05,
+      "loss": 0.5776,
+      "step": 5651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1617512640029863,
+      "learning_rate": 1.2453444654693099e-05,
+      "loss": 0.5895,
+      "step": 5652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1563366522409184,
+      "learning_rate": 1.245100870355057e-05,
+      "loss": 0.5467,
+      "step": 5653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1359980025307461,
+      "learning_rate": 1.2448572597663218e-05,
+      "loss": 0.5751,
+      "step": 5654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0829127057541117,
+      "learning_rate": 1.2446136337184847e-05,
+      "loss": 0.5796,
+      "step": 5655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1313118828133837,
+      "learning_rate": 1.2443699922269268e-05,
+      "loss": 0.5241,
+      "step": 5656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2017290513347676,
+      "learning_rate": 1.2441263353070305e-05,
+      "loss": 0.6428,
+      "step": 5657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2297537527692084,
+      "learning_rate": 1.243882662974179e-05,
+      "loss": 0.5594,
+      "step": 5658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2139172716311857,
+      "learning_rate": 1.2436389752437565e-05,
+      "loss": 0.5995,
+      "step": 5659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1963637970916101,
+      "learning_rate": 1.243395272131149e-05,
+      "loss": 0.5927,
+      "step": 5660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1619556998575276,
+      "learning_rate": 1.243151553651742e-05,
+      "loss": 0.5908,
+      "step": 5661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.136419512502045,
+      "learning_rate": 1.2429078198209227e-05,
+      "loss": 0.6033,
+      "step": 5662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2015794512243771,
+      "learning_rate": 1.2426640706540796e-05,
+      "loss": 0.6205,
+      "step": 5663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2196878101799054,
+      "learning_rate": 1.2424203061666018e-05,
+      "loss": 0.6196,
+      "step": 5664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1001492897510252,
+      "learning_rate": 1.2421765263738795e-05,
+      "loss": 0.5458,
+      "step": 5665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1485767409842995,
+      "learning_rate": 1.2419327312913034e-05,
+      "loss": 0.5601,
+      "step": 5666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2697973353963758,
+      "learning_rate": 1.2416889209342658e-05,
+      "loss": 0.6374,
+      "step": 5667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1694339505316522,
+      "learning_rate": 1.2414450953181598e-05,
+      "loss": 0.5717,
+      "step": 5668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2132129578735729,
+      "learning_rate": 1.2412012544583791e-05,
+      "loss": 0.5539,
+      "step": 5669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1065480924377782,
+      "learning_rate": 1.2409573983703189e-05,
+      "loss": 0.5327,
+      "step": 5670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2059210452271647,
+      "learning_rate": 1.2407135270693748e-05,
+      "loss": 0.5784,
+      "step": 5671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1985990930109207,
+      "learning_rate": 1.240469640570944e-05,
+      "loss": 0.5868,
+      "step": 5672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2559529650804542,
+      "learning_rate": 1.240225738890424e-05,
+      "loss": 0.6066,
+      "step": 5673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0678661819017123,
+      "learning_rate": 1.2399818220432136e-05,
+      "loss": 0.5225,
+      "step": 5674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1958938880147316,
+      "learning_rate": 1.239737890044713e-05,
+      "loss": 0.5535,
+      "step": 5675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1642319216742743,
+      "learning_rate": 1.2394939429103224e-05,
+      "loss": 0.5682,
+      "step": 5676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1415536314053027,
+      "learning_rate": 1.2392499806554433e-05,
+      "loss": 0.5845,
+      "step": 5677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1686232146174917,
+      "learning_rate": 1.2390060032954787e-05,
+      "loss": 0.5718,
+      "step": 5678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.214916950731231,
+      "learning_rate": 1.2387620108458318e-05,
+      "loss": 0.5272,
+      "step": 5679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1916406119550509,
+      "learning_rate": 1.2385180033219077e-05,
+      "loss": 0.5934,
+      "step": 5680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1582562850923244,
+      "learning_rate": 1.2382739807391113e-05,
+      "loss": 0.5916,
+      "step": 5681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1741218389848078,
+      "learning_rate": 1.238029943112849e-05,
+      "loss": 0.51,
+      "step": 5682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9802493575642978,
+      "learning_rate": 1.2377858904585284e-05,
+      "loss": 0.5269,
+      "step": 5683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.252618336217392,
+      "learning_rate": 1.237541822791558e-05,
+      "loss": 0.5678,
+      "step": 5684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0191928111788477,
+      "learning_rate": 1.2372977401273465e-05,
+      "loss": 0.535,
+      "step": 5685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1349284437704643,
+      "learning_rate": 1.2370536424813044e-05,
+      "loss": 0.5501,
+      "step": 5686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1382052739851187,
+      "learning_rate": 1.2368095298688428e-05,
+      "loss": 0.535,
+      "step": 5687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.134629259434978,
+      "learning_rate": 1.236565402305374e-05,
+      "loss": 0.5907,
+      "step": 5688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.217449668298233,
+      "learning_rate": 1.2363212598063103e-05,
+      "loss": 0.6254,
+      "step": 5689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2451542388788717,
+      "learning_rate": 1.2360771023870668e-05,
+      "loss": 0.6002,
+      "step": 5690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2742246289705255,
+      "learning_rate": 1.2358329300630576e-05,
+      "loss": 0.6187,
+      "step": 5691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1497961298975805,
+      "learning_rate": 1.2355887428496986e-05,
+      "loss": 0.5844,
+      "step": 5692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.109832991502645,
+      "learning_rate": 1.2353445407624071e-05,
+      "loss": 0.5787,
+      "step": 5693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0901601462002743,
+      "learning_rate": 1.2351003238166004e-05,
+      "loss": 0.5423,
+      "step": 5694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1664810146482412,
+      "learning_rate": 1.2348560920276973e-05,
+      "loss": 0.5297,
+      "step": 5695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1852418107975102,
+      "learning_rate": 1.2346118454111176e-05,
+      "loss": 0.5865,
+      "step": 5696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2308784892572304,
+      "learning_rate": 1.2343675839822813e-05,
+      "loss": 0.5934,
+      "step": 5697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1731898751304157,
+      "learning_rate": 1.2341233077566104e-05,
+      "loss": 0.5783,
+      "step": 5698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.158076416083538,
+      "learning_rate": 1.2338790167495272e-05,
+      "loss": 0.5773,
+      "step": 5699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0918072476516998,
+      "learning_rate": 1.2336347109764551e-05,
+      "loss": 0.5666,
+      "step": 5700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.172473907017691,
+      "learning_rate": 1.2333903904528182e-05,
+      "loss": 0.5694,
+      "step": 5701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1775588985056196,
+      "learning_rate": 1.2331460551940417e-05,
+      "loss": 0.5915,
+      "step": 5702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9588980703889227,
+      "learning_rate": 1.232901705215552e-05,
+      "loss": 0.5136,
+      "step": 5703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2558507844859523,
+      "learning_rate": 1.232657340532776e-05,
+      "loss": 0.599,
+      "step": 5704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.279749852711989,
+      "learning_rate": 1.2324129611611417e-05,
+      "loss": 0.6208,
+      "step": 5705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.218997245553889,
+      "learning_rate": 1.2321685671160784e-05,
+      "loss": 0.5686,
+      "step": 5706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0930544140002052,
+      "learning_rate": 1.231924158413015e-05,
+      "loss": 0.5025,
+      "step": 5707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2123553927207038,
+      "learning_rate": 1.2316797350673834e-05,
+      "loss": 0.5576,
+      "step": 5708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.299455534185615,
+      "learning_rate": 1.2314352970946146e-05,
+      "loss": 0.5765,
+      "step": 5709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1586965503731155,
+      "learning_rate": 1.2311908445101414e-05,
+      "loss": 0.6,
+      "step": 5710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.284343845332993,
+      "learning_rate": 1.2309463773293977e-05,
+      "loss": 0.6017,
+      "step": 5711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0505608536689617,
+      "learning_rate": 1.2307018955678174e-05,
+      "loss": 0.5418,
+      "step": 5712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1649264003769886,
+      "learning_rate": 1.2304573992408363e-05,
+      "loss": 0.5433,
+      "step": 5713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1021096784667377,
+      "learning_rate": 1.2302128883638904e-05,
+      "loss": 0.5567,
+      "step": 5714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1644419621262367,
+      "learning_rate": 1.2299683629524175e-05,
+      "loss": 0.5814,
+      "step": 5715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3517347457586624,
+      "learning_rate": 1.2297238230218551e-05,
+      "loss": 0.6268,
+      "step": 5716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2117459951420755,
+      "learning_rate": 1.2294792685876424e-05,
+      "loss": 0.5515,
+      "step": 5717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1427577733368208,
+      "learning_rate": 1.2292346996652198e-05,
+      "loss": 0.5639,
+      "step": 5718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1971142223518676,
+      "learning_rate": 1.2289901162700276e-05,
+      "loss": 0.5631,
+      "step": 5719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2168086802979947,
+      "learning_rate": 1.228745518417508e-05,
+      "loss": 0.622,
+      "step": 5720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2120541254176294,
+      "learning_rate": 1.228500906123104e-05,
+      "loss": 0.5595,
+      "step": 5721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2924462498716816,
+      "learning_rate": 1.2282562794022586e-05,
+      "loss": 0.6143,
+      "step": 5722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2596763401772133,
+      "learning_rate": 1.2280116382704166e-05,
+      "loss": 0.5961,
+      "step": 5723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3843915296820637,
+      "learning_rate": 1.2277669827430234e-05,
+      "loss": 0.6031,
+      "step": 5724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.093212976770467,
+      "learning_rate": 1.2275223128355258e-05,
+      "loss": 0.5511,
+      "step": 5725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1682795490631581,
+      "learning_rate": 1.2272776285633708e-05,
+      "loss": 0.5524,
+      "step": 5726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2210305713166165,
+      "learning_rate": 1.2270329299420061e-05,
+      "loss": 0.6075,
+      "step": 5727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4010380694072542,
+      "learning_rate": 1.2267882169868813e-05,
+      "loss": 0.5523,
+      "step": 5728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1338893113350184,
+      "learning_rate": 1.2265434897134462e-05,
+      "loss": 0.5625,
+      "step": 5729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1550434622080412,
+      "learning_rate": 1.2262987481371523e-05,
+      "loss": 0.5506,
+      "step": 5730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0366448405524578,
+      "learning_rate": 1.2260539922734505e-05,
+      "loss": 0.5239,
+      "step": 5731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.378430119513949,
+      "learning_rate": 1.2258092221377938e-05,
+      "loss": 0.6496,
+      "step": 5732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2256839302675868,
+      "learning_rate": 1.2255644377456357e-05,
+      "loss": 0.5612,
+      "step": 5733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.279751902020504,
+      "learning_rate": 1.2253196391124313e-05,
+      "loss": 0.6126,
+      "step": 5734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1559703591846446,
+      "learning_rate": 1.2250748262536357e-05,
+      "loss": 0.5656,
+      "step": 5735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1631600895200256,
+      "learning_rate": 1.2248299991847048e-05,
+      "loss": 0.4909,
+      "step": 5736
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1719034827267767,
+      "learning_rate": 1.2245851579210958e-05,
+      "loss": 0.5376,
+      "step": 5737
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1778271387111043,
+      "learning_rate": 1.2243403024782675e-05,
+      "loss": 0.5458,
+      "step": 5738
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2291002201719787,
+      "learning_rate": 1.2240954328716783e-05,
+      "loss": 0.5324,
+      "step": 5739
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1940587767857695,
+      "learning_rate": 1.2238505491167884e-05,
+      "loss": 0.5983,
+      "step": 5740
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1452531848655838,
+      "learning_rate": 1.2236056512290584e-05,
+      "loss": 0.5081,
+      "step": 5741
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1813157825217495,
+      "learning_rate": 1.2233607392239497e-05,
+      "loss": 0.5692,
+      "step": 5742
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.171500082441743,
+      "learning_rate": 1.2231158131169251e-05,
+      "loss": 0.5278,
+      "step": 5743
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1124261617266489,
+      "learning_rate": 1.2228708729234487e-05,
+      "loss": 0.5186,
+      "step": 5744
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0741828496309291,
+      "learning_rate": 1.222625918658984e-05,
+      "loss": 0.5729,
+      "step": 5745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3032568665891962,
+      "learning_rate": 1.2223809503389962e-05,
+      "loss": 0.5913,
+      "step": 5746
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.172448285066508,
+      "learning_rate": 1.2221359679789518e-05,
+      "loss": 0.6391,
+      "step": 5747
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1912769184893888,
+      "learning_rate": 1.2218909715943174e-05,
+      "loss": 0.5524,
+      "step": 5748
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.190448413479408,
+      "learning_rate": 1.2216459612005616e-05,
+      "loss": 0.5552,
+      "step": 5749
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1883309117560483,
+      "learning_rate": 1.2214009368131525e-05,
+      "loss": 0.5022,
+      "step": 5750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.273073905103042,
+      "learning_rate": 1.22115589844756e-05,
+      "loss": 0.5961,
+      "step": 5751
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3431829098678116,
+      "learning_rate": 1.2209108461192546e-05,
+      "loss": 0.5901,
+      "step": 5752
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.135715736874924,
+      "learning_rate": 1.2206657798437078e-05,
+      "loss": 0.541,
+      "step": 5753
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1776193338008218,
+      "learning_rate": 1.2204206996363918e-05,
+      "loss": 0.5781,
+      "step": 5754
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2144509321299104,
+      "learning_rate": 1.2201756055127798e-05,
+      "loss": 0.5979,
+      "step": 5755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0928416158426182,
+      "learning_rate": 1.219930497488346e-05,
+      "loss": 0.573,
+      "step": 5756
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1879574246050468,
+      "learning_rate": 1.2196853755785649e-05,
+      "loss": 0.5774,
+      "step": 5757
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2330869400639546,
+      "learning_rate": 1.2194402397989128e-05,
+      "loss": 0.5453,
+      "step": 5758
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1897570091885952,
+      "learning_rate": 1.2191950901648664e-05,
+      "loss": 0.527,
+      "step": 5759
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.080463637245422,
+      "learning_rate": 1.2189499266919028e-05,
+      "loss": 0.5033,
+      "step": 5760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1655886074245116,
+      "learning_rate": 1.2187047493955006e-05,
+      "loss": 0.5931,
+      "step": 5761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.174902197643059,
+      "learning_rate": 1.2184595582911394e-05,
+      "loss": 0.5324,
+      "step": 5762
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2115452379635694,
+      "learning_rate": 1.218214353394299e-05,
+      "loss": 0.5163,
+      "step": 5763
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1526727756237607,
+      "learning_rate": 1.2179691347204609e-05,
+      "loss": 0.4868,
+      "step": 5764
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2146941457791842,
+      "learning_rate": 1.2177239022851068e-05,
+      "loss": 0.5646,
+      "step": 5765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1720135416153417,
+      "learning_rate": 1.2174786561037191e-05,
+      "loss": 0.5774,
+      "step": 5766
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2000503390403146,
+      "learning_rate": 1.2172333961917819e-05,
+      "loss": 0.5609,
+      "step": 5767
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1779116470253348,
+      "learning_rate": 1.2169881225647797e-05,
+      "loss": 0.5655,
+      "step": 5768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1407609100890166,
+      "learning_rate": 1.2167428352381977e-05,
+      "loss": 0.5076,
+      "step": 5769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0374528276680997,
+      "learning_rate": 1.2164975342275227e-05,
+      "loss": 0.514,
+      "step": 5770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1877798955226615,
+      "learning_rate": 1.2162522195482408e-05,
+      "loss": 0.5438,
+      "step": 5771
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1487102573896208,
+      "learning_rate": 1.2160068912158408e-05,
+      "loss": 0.544,
+      "step": 5772
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2242168180624513,
+      "learning_rate": 1.2157615492458113e-05,
+      "loss": 0.5536,
+      "step": 5773
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.101818082121585,
+      "learning_rate": 1.2155161936536422e-05,
+      "loss": 0.5893,
+      "step": 5774
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.083124538009906,
+      "learning_rate": 1.2152708244548237e-05,
+      "loss": 0.5432,
+      "step": 5775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1948722577151858,
+      "learning_rate": 1.2150254416648473e-05,
+      "loss": 0.541,
+      "step": 5776
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2256619980755752,
+      "learning_rate": 1.2147800452992055e-05,
+      "loss": 0.5762,
+      "step": 5777
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1837845953747075,
+      "learning_rate": 1.2145346353733913e-05,
+      "loss": 0.5393,
+      "step": 5778
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1906830140245028,
+      "learning_rate": 1.2142892119028988e-05,
+      "loss": 0.5614,
+      "step": 5779
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1461272065129666,
+      "learning_rate": 1.2140437749032227e-05,
+      "loss": 0.5559,
+      "step": 5780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1644388396965029,
+      "learning_rate": 1.2137983243898585e-05,
+      "loss": 0.565,
+      "step": 5781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2571805705259522,
+      "learning_rate": 1.2135528603783036e-05,
+      "loss": 0.5691,
+      "step": 5782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0542582450763196,
+      "learning_rate": 1.2133073828840543e-05,
+      "loss": 0.5131,
+      "step": 5783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1099808341596908,
+      "learning_rate": 1.2130618919226099e-05,
+      "loss": 0.5439,
+      "step": 5784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2504267917635397,
+      "learning_rate": 1.2128163875094687e-05,
+      "loss": 0.614,
+      "step": 5785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2536251430391578,
+      "learning_rate": 1.2125708696601309e-05,
+      "loss": 0.64,
+      "step": 5786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2027608146871105,
+      "learning_rate": 1.2123253383900974e-05,
+      "loss": 0.579,
+      "step": 5787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2219459005795537,
+      "learning_rate": 1.2120797937148699e-05,
+      "loss": 0.608,
+      "step": 5788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1487539984101123,
+      "learning_rate": 1.2118342356499508e-05,
+      "loss": 0.5553,
+      "step": 5789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0749243620943467,
+      "learning_rate": 1.2115886642108437e-05,
+      "loss": 0.5285,
+      "step": 5790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0901008767511062,
+      "learning_rate": 1.2113430794130522e-05,
+      "loss": 0.5143,
+      "step": 5791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.109954843633605,
+      "learning_rate": 1.2110974812720819e-05,
+      "loss": 0.5127,
+      "step": 5792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1113190608668968,
+      "learning_rate": 1.2108518698034384e-05,
+      "loss": 0.5305,
+      "step": 5793
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1599832527826188,
+      "learning_rate": 1.2106062450226287e-05,
+      "loss": 0.5953,
+      "step": 5794
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1504900489419883,
+      "learning_rate": 1.2103606069451601e-05,
+      "loss": 0.5835,
+      "step": 5795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2313824848601616,
+      "learning_rate": 1.210114955586541e-05,
+      "loss": 0.6057,
+      "step": 5796
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2464898893831475,
+      "learning_rate": 1.2098692909622808e-05,
+      "loss": 0.5481,
+      "step": 5797
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2924023911974978,
+      "learning_rate": 1.2096236130878894e-05,
+      "loss": 0.5437,
+      "step": 5798
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1411081623293935,
+      "learning_rate": 1.2093779219788777e-05,
+      "loss": 0.5064,
+      "step": 5799
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.062769631225352,
+      "learning_rate": 1.2091322176507579e-05,
+      "loss": 0.4804,
+      "step": 5800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0799403180303564,
+      "learning_rate": 1.2088865001190418e-05,
+      "loss": 0.5354,
+      "step": 5801
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2463022374537462,
+      "learning_rate": 1.2086407693992434e-05,
+      "loss": 0.6065,
+      "step": 5802
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.106121396278204,
+      "learning_rate": 1.2083950255068766e-05,
+      "loss": 0.5653,
+      "step": 5803
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1750484476855685,
+      "learning_rate": 1.208149268457457e-05,
+      "loss": 0.5781,
+      "step": 5804
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.173305147430331,
+      "learning_rate": 1.2079034982665001e-05,
+      "loss": 0.5533,
+      "step": 5805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1877700097147526,
+      "learning_rate": 1.2076577149495226e-05,
+      "loss": 0.575,
+      "step": 5806
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.226780185698904,
+      "learning_rate": 1.207411918522042e-05,
+      "loss": 0.5951,
+      "step": 5807
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1669546464444174,
+      "learning_rate": 1.2071661089995772e-05,
+      "loss": 0.5916,
+      "step": 5808
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.147030493981939,
+      "learning_rate": 1.2069202863976471e-05,
+      "loss": 0.5472,
+      "step": 5809
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1124389138756328,
+      "learning_rate": 1.2066744507317718e-05,
+      "loss": 0.566,
+      "step": 5810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2951715151052539,
+      "learning_rate": 1.2064286020174718e-05,
+      "loss": 0.5823,
+      "step": 5811
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2010264238822317,
+      "learning_rate": 1.2061827402702691e-05,
+      "loss": 0.5236,
+      "step": 5812
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1301946448743838,
+      "learning_rate": 1.2059368655056864e-05,
+      "loss": 0.5331,
+      "step": 5813
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1961510403477786,
+      "learning_rate": 1.2056909777392471e-05,
+      "loss": 0.6108,
+      "step": 5814
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.18671762894558,
+      "learning_rate": 1.2054450769864752e-05,
+      "loss": 0.5675,
+      "step": 5815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1760007330249915,
+      "learning_rate": 1.2051991632628952e-05,
+      "loss": 0.5775,
+      "step": 5816
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.16608463255676,
+      "learning_rate": 1.2049532365840333e-05,
+      "loss": 0.6161,
+      "step": 5817
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1621810249872406,
+      "learning_rate": 1.2047072969654165e-05,
+      "loss": 0.5382,
+      "step": 5818
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1802851192660717,
+      "learning_rate": 1.204461344422572e-05,
+      "loss": 0.5502,
+      "step": 5819
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1400551613629568,
+      "learning_rate": 1.2042153789710278e-05,
+      "loss": 0.5125,
+      "step": 5820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1618702361946471,
+      "learning_rate": 1.2039694006263129e-05,
+      "loss": 0.5481,
+      "step": 5821
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1806254416081747,
+      "learning_rate": 1.2037234094039573e-05,
+      "loss": 0.5591,
+      "step": 5822
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2368320203564345,
+      "learning_rate": 1.2034774053194922e-05,
+      "loss": 0.5767,
+      "step": 5823
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1410431813820914,
+      "learning_rate": 1.2032313883884485e-05,
+      "loss": 0.5382,
+      "step": 5824
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.222155678748634,
+      "learning_rate": 1.202985358626359e-05,
+      "loss": 0.5459,
+      "step": 5825
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.202775978887758,
+      "learning_rate": 1.2027393160487561e-05,
+      "loss": 0.5589,
+      "step": 5826
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1627100826474233,
+      "learning_rate": 1.2024932606711741e-05,
+      "loss": 0.5983,
+      "step": 5827
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1437455432575716,
+      "learning_rate": 1.2022471925091483e-05,
+      "loss": 0.5652,
+      "step": 5828
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.13861978977748,
+      "learning_rate": 1.2020011115782135e-05,
+      "loss": 0.5256,
+      "step": 5829
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0433482076317901,
+      "learning_rate": 1.2017550178939064e-05,
+      "loss": 0.5304,
+      "step": 5830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1142188156485253,
+      "learning_rate": 1.2015089114717642e-05,
+      "loss": 0.5115,
+      "step": 5831
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1799556099090744,
+      "learning_rate": 1.2012627923273246e-05,
+      "loss": 0.5378,
+      "step": 5832
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2779304020486262,
+      "learning_rate": 1.2010166604761266e-05,
+      "loss": 0.5505,
+      "step": 5833
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2344693015087467,
+      "learning_rate": 1.20077051593371e-05,
+      "loss": 0.5774,
+      "step": 5834
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1387612780484628,
+      "learning_rate": 1.2005243587156143e-05,
+      "loss": 0.5335,
+      "step": 5835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1495471186992472,
+      "learning_rate": 1.2002781888373818e-05,
+      "loss": 0.5972,
+      "step": 5836
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1760604374854247,
+      "learning_rate": 1.2000320063145536e-05,
+      "loss": 0.6117,
+      "step": 5837
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1046612489594707,
+      "learning_rate": 1.199785811162673e-05,
+      "loss": 0.5099,
+      "step": 5838
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1250570070869994,
+      "learning_rate": 1.1995396033972834e-05,
+      "loss": 0.5631,
+      "step": 5839
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.110835765789218,
+      "learning_rate": 1.1992933830339288e-05,
+      "loss": 0.502,
+      "step": 5840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1313335893634813,
+      "learning_rate": 1.1990471500881551e-05,
+      "loss": 0.593,
+      "step": 5841
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1774903609251823,
+      "learning_rate": 1.1988009045755077e-05,
+      "loss": 0.5985,
+      "step": 5842
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2243414042655396,
+      "learning_rate": 1.1985546465115336e-05,
+      "loss": 0.5613,
+      "step": 5843
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1738508733236663,
+      "learning_rate": 1.1983083759117804e-05,
+      "loss": 0.5925,
+      "step": 5844
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0072078813318732,
+      "learning_rate": 1.1980620927917958e-05,
+      "loss": 0.5195,
+      "step": 5845
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1427341452470745,
+      "learning_rate": 1.19781579716713e-05,
+      "loss": 0.5703,
+      "step": 5846
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3303544573880208,
+      "learning_rate": 1.1975694890533318e-05,
+      "loss": 0.5964,
+      "step": 5847
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2233942504097892,
+      "learning_rate": 1.1973231684659527e-05,
+      "loss": 0.6063,
+      "step": 5848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2731503121842511,
+      "learning_rate": 1.1970768354205443e-05,
+      "loss": 0.5583,
+      "step": 5849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0754297816765037,
+      "learning_rate": 1.1968304899326577e-05,
+      "loss": 0.5909,
+      "step": 5850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1399600559899563,
+      "learning_rate": 1.1965841320178473e-05,
+      "loss": 0.5293,
+      "step": 5851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2077981432299578,
+      "learning_rate": 1.1963377616916662e-05,
+      "loss": 0.5339,
+      "step": 5852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2131592089208965,
+      "learning_rate": 1.1960913789696694e-05,
+      "loss": 0.5758,
+      "step": 5853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.215401769700864,
+      "learning_rate": 1.1958449838674122e-05,
+      "loss": 0.5389,
+      "step": 5854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1799737444074103,
+      "learning_rate": 1.1955985764004504e-05,
+      "loss": 0.5464,
+      "step": 5855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2863515843054454,
+      "learning_rate": 1.1953521565843415e-05,
+      "loss": 0.6042,
+      "step": 5856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.204111561729822,
+      "learning_rate": 1.195105724434643e-05,
+      "loss": 0.5337,
+      "step": 5857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1542259233230394,
+      "learning_rate": 1.1948592799669136e-05,
+      "loss": 0.5513,
+      "step": 5858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0744879367980187,
+      "learning_rate": 1.1946128231967127e-05,
+      "loss": 0.5253,
+      "step": 5859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0442689941592997,
+      "learning_rate": 1.1943663541395998e-05,
+      "loss": 0.5464,
+      "step": 5860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1750569187744138,
+      "learning_rate": 1.1941198728111364e-05,
+      "loss": 0.571,
+      "step": 5861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1717132456724335,
+      "learning_rate": 1.1938733792268837e-05,
+      "loss": 0.5556,
+      "step": 5862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1851471630897263,
+      "learning_rate": 1.1936268734024048e-05,
+      "loss": 0.5668,
+      "step": 5863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1747150339368848,
+      "learning_rate": 1.1933803553532622e-05,
+      "loss": 0.5544,
+      "step": 5864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1452167007744813,
+      "learning_rate": 1.1931338250950197e-05,
+      "loss": 0.5183,
+      "step": 5865
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.194493181259474,
+      "learning_rate": 1.1928872826432427e-05,
+      "loss": 0.5388,
+      "step": 5866
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1237559327508897,
+      "learning_rate": 1.1926407280134962e-05,
+      "loss": 0.5561,
+      "step": 5867
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3663842593692355,
+      "learning_rate": 1.1923941612213468e-05,
+      "loss": 0.5684,
+      "step": 5868
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2856317817238363,
+      "learning_rate": 1.1921475822823613e-05,
+      "loss": 0.6646,
+      "step": 5869
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0807219478258843,
+      "learning_rate": 1.1919009912121075e-05,
+      "loss": 0.5468,
+      "step": 5870
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1001585542676102,
+      "learning_rate": 1.1916543880261541e-05,
+      "loss": 0.5462,
+      "step": 5871
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.210953103241838,
+      "learning_rate": 1.1914077727400706e-05,
+      "loss": 0.5531,
+      "step": 5872
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1054588290466927,
+      "learning_rate": 1.1911611453694267e-05,
+      "loss": 0.5374,
+      "step": 5873
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.14059197364965,
+      "learning_rate": 1.1909145059297935e-05,
+      "loss": 0.527,
+      "step": 5874
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1699562131399075,
+      "learning_rate": 1.1906678544367423e-05,
+      "loss": 0.5134,
+      "step": 5875
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2421120194828592,
+      "learning_rate": 1.190421190905846e-05,
+      "loss": 0.5647,
+      "step": 5876
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2952461124428827,
+      "learning_rate": 1.1901745153526773e-05,
+      "loss": 0.5676,
+      "step": 5877
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2217136442373169,
+      "learning_rate": 1.1899278277928103e-05,
+      "loss": 0.594,
+      "step": 5878
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1674616296413667,
+      "learning_rate": 1.1896811282418199e-05,
+      "loss": 0.5698,
+      "step": 5879
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.213094598938523,
+      "learning_rate": 1.1894344167152809e-05,
+      "loss": 0.6098,
+      "step": 5880
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.238448026479051,
+      "learning_rate": 1.1891876932287701e-05,
+      "loss": 0.6178,
+      "step": 5881
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2836003494325223,
+      "learning_rate": 1.1889409577978639e-05,
+      "loss": 0.6002,
+      "step": 5882
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1942288841247801,
+      "learning_rate": 1.1886942104381403e-05,
+      "loss": 0.5484,
+      "step": 5883
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3050263929803976,
+      "learning_rate": 1.1884474511651778e-05,
+      "loss": 0.5789,
+      "step": 5884
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.208662739062639,
+      "learning_rate": 1.1882006799945551e-05,
+      "loss": 0.6153,
+      "step": 5885
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0800346379553527,
+      "learning_rate": 1.1879538969418526e-05,
+      "loss": 0.568,
+      "step": 5886
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1419719424801138,
+      "learning_rate": 1.1877071020226512e-05,
+      "loss": 0.5483,
+      "step": 5887
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2478569733996008,
+      "learning_rate": 1.1874602952525317e-05,
+      "loss": 0.5996,
+      "step": 5888
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3204928844856738,
+      "learning_rate": 1.1872134766470769e-05,
+      "loss": 0.6373,
+      "step": 5889
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1594939219721387,
+      "learning_rate": 1.1869666462218693e-05,
+      "loss": 0.5458,
+      "step": 5890
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0793814593874516,
+      "learning_rate": 1.1867198039924923e-05,
+      "loss": 0.516,
+      "step": 5891
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1567645088079015,
+      "learning_rate": 1.1864729499745312e-05,
+      "loss": 0.5673,
+      "step": 5892
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0312249729703138,
+      "learning_rate": 1.1862260841835706e-05,
+      "loss": 0.5082,
+      "step": 5893
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.086463690964292,
+      "learning_rate": 1.1859792066351964e-05,
+      "loss": 0.5152,
+      "step": 5894
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.121048344878929,
+      "learning_rate": 1.1857323173449956e-05,
+      "loss": 0.5682,
+      "step": 5895
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3202165196310283,
+      "learning_rate": 1.1854854163285548e-05,
+      "loss": 0.6055,
+      "step": 5896
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.186041136624872,
+      "learning_rate": 1.185238503601463e-05,
+      "loss": 0.5791,
+      "step": 5897
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.22254216938579,
+      "learning_rate": 1.1849915791793091e-05,
+      "loss": 0.5845,
+      "step": 5898
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1606132044113173,
+      "learning_rate": 1.1847446430776822e-05,
+      "loss": 0.5692,
+      "step": 5899
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1930835361379735,
+      "learning_rate": 1.1844976953121725e-05,
+      "loss": 0.5913,
+      "step": 5900
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1986993416592568,
+      "learning_rate": 1.1842507358983715e-05,
+      "loss": 0.6328,
+      "step": 5901
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1882834610677058,
+      "learning_rate": 1.1840037648518712e-05,
+      "loss": 0.5788,
+      "step": 5902
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1837200940875108,
+      "learning_rate": 1.1837567821882638e-05,
+      "loss": 0.5791,
+      "step": 5903
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1003249772070514,
+      "learning_rate": 1.1835097879231427e-05,
+      "loss": 0.5656,
+      "step": 5904
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2701610235574328,
+      "learning_rate": 1.1832627820721017e-05,
+      "loss": 0.5365,
+      "step": 5905
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.143477804594884,
+      "learning_rate": 1.1830157646507358e-05,
+      "loss": 0.5089,
+      "step": 5906
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2677134484235162,
+      "learning_rate": 1.1827687356746406e-05,
+      "loss": 0.5629,
+      "step": 5907
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1531484089778088,
+      "learning_rate": 1.182521695159412e-05,
+      "loss": 0.539,
+      "step": 5908
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1058788280332412,
+      "learning_rate": 1.1822746431206473e-05,
+      "loss": 0.5695,
+      "step": 5909
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1808895025217803,
+      "learning_rate": 1.1820275795739438e-05,
+      "loss": 0.5691,
+      "step": 5910
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2212088307964457,
+      "learning_rate": 1.1817805045349e-05,
+      "loss": 0.6193,
+      "step": 5911
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1192837768393247,
+      "learning_rate": 1.1815334180191153e-05,
+      "loss": 0.5424,
+      "step": 5912
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2534083153145257,
+      "learning_rate": 1.1812863200421894e-05,
+      "loss": 0.5966,
+      "step": 5913
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2925274144701855,
+      "learning_rate": 1.1810392106197224e-05,
+      "loss": 0.6072,
+      "step": 5914
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1839332723082427,
+      "learning_rate": 1.1807920897673162e-05,
+      "loss": 0.55,
+      "step": 5915
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1519622688538522,
+      "learning_rate": 1.1805449575005726e-05,
+      "loss": 0.5678,
+      "step": 5916
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1629580179970105,
+      "learning_rate": 1.1802978138350945e-05,
+      "loss": 0.5458,
+      "step": 5917
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1714943839902012,
+      "learning_rate": 1.1800506587864851e-05,
+      "loss": 0.5418,
+      "step": 5918
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.243544838628027,
+      "learning_rate": 1.1798034923703486e-05,
+      "loss": 0.6272,
+      "step": 5919
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1981541881876347,
+      "learning_rate": 1.17955631460229e-05,
+      "loss": 0.519,
+      "step": 5920
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1811447237927264,
+      "learning_rate": 1.1793091254979148e-05,
+      "loss": 0.5514,
+      "step": 5921
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1235853413593588,
+      "learning_rate": 1.1790619250728295e-05,
+      "loss": 0.5534,
+      "step": 5922
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1624878851966867,
+      "learning_rate": 1.178814713342641e-05,
+      "loss": 0.5502,
+      "step": 5923
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.114220260000697,
+      "learning_rate": 1.1785674903229572e-05,
+      "loss": 0.55,
+      "step": 5924
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2299317171945692,
+      "learning_rate": 1.1783202560293863e-05,
+      "loss": 0.5908,
+      "step": 5925
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.234587301032679,
+      "learning_rate": 1.1780730104775374e-05,
+      "loss": 0.5998,
+      "step": 5926
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0639864117141906,
+      "learning_rate": 1.1778257536830211e-05,
+      "loss": 0.5353,
+      "step": 5927
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1595802289999118,
+      "learning_rate": 1.1775784856614473e-05,
+      "loss": 0.5223,
+      "step": 5928
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.125575077574416,
+      "learning_rate": 1.1773312064284275e-05,
+      "loss": 0.5552,
+      "step": 5929
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1473486273162008,
+      "learning_rate": 1.1770839159995738e-05,
+      "loss": 0.5426,
+      "step": 5930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.132527966775059,
+      "learning_rate": 1.1768366143904986e-05,
+      "loss": 0.5297,
+      "step": 5931
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0903657599637857,
+      "learning_rate": 1.1765893016168158e-05,
+      "loss": 0.5232,
+      "step": 5932
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1708886636893996,
+      "learning_rate": 1.1763419776941395e-05,
+      "loss": 0.616,
+      "step": 5933
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2104614798850823,
+      "learning_rate": 1.1760946426380838e-05,
+      "loss": 0.5964,
+      "step": 5934
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1134263328915142,
+      "learning_rate": 1.1758472964642651e-05,
+      "loss": 0.5122,
+      "step": 5935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1541368919048283,
+      "learning_rate": 1.1755999391882993e-05,
+      "loss": 0.5348,
+      "step": 5936
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1571005708378703,
+      "learning_rate": 1.1753525708258034e-05,
+      "loss": 0.5326,
+      "step": 5937
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1534524013452785,
+      "learning_rate": 1.175105191392395e-05,
+      "loss": 0.6112,
+      "step": 5938
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.207627183285773,
+      "learning_rate": 1.1748578009036925e-05,
+      "loss": 0.5764,
+      "step": 5939
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0799936327040143,
+      "learning_rate": 1.1746103993753146e-05,
+      "loss": 0.4969,
+      "step": 5940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0990050910054272,
+      "learning_rate": 1.1743629868228815e-05,
+      "loss": 0.5508,
+      "step": 5941
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0899727584339496,
+      "learning_rate": 1.1741155632620135e-05,
+      "loss": 0.5599,
+      "step": 5942
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.192224390501563,
+      "learning_rate": 1.1738681287083318e-05,
+      "loss": 0.5345,
+      "step": 5943
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0077912444536987,
+      "learning_rate": 1.1736206831774576e-05,
+      "loss": 0.5608,
+      "step": 5944
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9353286074428118,
+      "learning_rate": 1.1733732266850144e-05,
+      "loss": 0.4908,
+      "step": 5945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0984879461780905,
+      "learning_rate": 1.1731257592466248e-05,
+      "loss": 0.555,
+      "step": 5946
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0896925735902507,
+      "learning_rate": 1.1728782808779126e-05,
+      "loss": 0.5817,
+      "step": 5947
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1967637471018489,
+      "learning_rate": 1.172630791594503e-05,
+      "loss": 0.5933,
+      "step": 5948
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0948715454249143,
+      "learning_rate": 1.1723832914120203e-05,
+      "loss": 0.5489,
+      "step": 5949
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.12736728403082,
+      "learning_rate": 1.1721357803460915e-05,
+      "loss": 0.5393,
+      "step": 5950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1518496213653773,
+      "learning_rate": 1.1718882584123425e-05,
+      "loss": 0.5458,
+      "step": 5951
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1689298114051585,
+      "learning_rate": 1.1716407256264014e-05,
+      "loss": 0.5602,
+      "step": 5952
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1658066349821727,
+      "learning_rate": 1.1713931820038952e-05,
+      "loss": 0.5317,
+      "step": 5953
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1068523900519198,
+      "learning_rate": 1.1711456275604534e-05,
+      "loss": 0.595,
+      "step": 5954
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0445035578410105,
+      "learning_rate": 1.170898062311705e-05,
+      "loss": 0.5791,
+      "step": 5955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0967450143680333,
+      "learning_rate": 1.1706504862732801e-05,
+      "loss": 0.527,
+      "step": 5956
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2187901514726485,
+      "learning_rate": 1.17040289946081e-05,
+      "loss": 0.6348,
+      "step": 5957
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1074112718400848,
+      "learning_rate": 1.1701553018899255e-05,
+      "loss": 0.4888,
+      "step": 5958
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1410849179622504,
+      "learning_rate": 1.1699076935762585e-05,
+      "loss": 0.5592,
+      "step": 5959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.206214653248334,
+      "learning_rate": 1.1696600745354427e-05,
+      "loss": 0.5846,
+      "step": 5960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1250317357143491,
+      "learning_rate": 1.1694124447831108e-05,
+      "loss": 0.5303,
+      "step": 5961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.008943676986175,
+      "learning_rate": 1.1691648043348972e-05,
+      "loss": 0.4977,
+      "step": 5962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1748748022704938,
+      "learning_rate": 1.1689171532064371e-05,
+      "loss": 0.5849,
+      "step": 5963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.223466988277587,
+      "learning_rate": 1.1686694914133652e-05,
+      "loss": 0.5684,
+      "step": 5964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.276509750565055,
+      "learning_rate": 1.1684218189713183e-05,
+      "loss": 0.5751,
+      "step": 5965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2072021955898635,
+      "learning_rate": 1.1681741358959328e-05,
+      "loss": 0.5492,
+      "step": 5966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2206299294461564,
+      "learning_rate": 1.1679264422028469e-05,
+      "loss": 0.6172,
+      "step": 5967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2630238588696714,
+      "learning_rate": 1.167678737907698e-05,
+      "loss": 0.5606,
+      "step": 5968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0384634485254356,
+      "learning_rate": 1.1674310230261251e-05,
+      "loss": 0.5534,
+      "step": 5969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0826442373090766,
+      "learning_rate": 1.167183297573768e-05,
+      "loss": 0.5247,
+      "step": 5970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.135455501253906,
+      "learning_rate": 1.166935561566267e-05,
+      "loss": 0.5342,
+      "step": 5971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1690433113748349,
+      "learning_rate": 1.1666878150192626e-05,
+      "loss": 0.5679,
+      "step": 5972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1313390686209732,
+      "learning_rate": 1.1664400579483965e-05,
+      "loss": 0.5665,
+      "step": 5973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1735206747228935,
+      "learning_rate": 1.1661922903693107e-05,
+      "loss": 0.5417,
+      "step": 5974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1505834032586726,
+      "learning_rate": 1.165944512297648e-05,
+      "loss": 0.5898,
+      "step": 5975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0794256905921384,
+      "learning_rate": 1.1656967237490524e-05,
+      "loss": 0.5122,
+      "step": 5976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1225912480128515,
+      "learning_rate": 1.1654489247391678e-05,
+      "loss": 0.5448,
+      "step": 5977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2004672332956667,
+      "learning_rate": 1.1652011152836388e-05,
+      "loss": 0.5809,
+      "step": 5978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1600084306969605,
+      "learning_rate": 1.1649532953981111e-05,
+      "loss": 0.6212,
+      "step": 5979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1847955123296074,
+      "learning_rate": 1.1647054650982306e-05,
+      "loss": 0.575,
+      "step": 5980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.089012512636066,
+      "learning_rate": 1.1644576243996446e-05,
+      "loss": 0.5012,
+      "step": 5981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2342043891531973,
+      "learning_rate": 1.1642097733180003e-05,
+      "loss": 0.5971,
+      "step": 5982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2085579411257474,
+      "learning_rate": 1.1639619118689456e-05,
+      "loss": 0.5863,
+      "step": 5983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1374002622537998,
+      "learning_rate": 1.1637140400681296e-05,
+      "loss": 0.5664,
+      "step": 5984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1734640410529102,
+      "learning_rate": 1.1634661579312012e-05,
+      "loss": 0.5435,
+      "step": 5985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1797157435794052,
+      "learning_rate": 1.1632182654738116e-05,
+      "loss": 0.5925,
+      "step": 5986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2278237145567825,
+      "learning_rate": 1.1629703627116104e-05,
+      "loss": 0.6195,
+      "step": 5987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0704874048930837,
+      "learning_rate": 1.1627224496602496e-05,
+      "loss": 0.5312,
+      "step": 5988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2364541421940813,
+      "learning_rate": 1.1624745263353808e-05,
+      "loss": 0.6087,
+      "step": 5989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1730046231958415,
+      "learning_rate": 1.1622265927526566e-05,
+      "loss": 0.5496,
+      "step": 5990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.035689594416083,
+      "learning_rate": 1.1619786489277312e-05,
+      "loss": 0.5185,
+      "step": 5991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1143388509744039,
+      "learning_rate": 1.1617306948762576e-05,
+      "loss": 0.5245,
+      "step": 5992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2708955916283018,
+      "learning_rate": 1.1614827306138912e-05,
+      "loss": 0.5783,
+      "step": 5993
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.113705631147262,
+      "learning_rate": 1.1612347561562865e-05,
+      "loss": 0.5612,
+      "step": 5994
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1133749403764557,
+      "learning_rate": 1.1609867715190997e-05,
+      "loss": 0.5667,
+      "step": 5995
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1728270160479073,
+      "learning_rate": 1.1607387767179881e-05,
+      "loss": 0.5273,
+      "step": 5996
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.132064361825146,
+      "learning_rate": 1.1604907717686075e-05,
+      "loss": 0.5189,
+      "step": 5997
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1384956655373737,
+      "learning_rate": 1.1602427566866166e-05,
+      "loss": 0.5398,
+      "step": 5998
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2144947102550459,
+      "learning_rate": 1.1599947314876738e-05,
+      "loss": 0.5539,
+      "step": 5999
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.119428294759224,
+      "learning_rate": 1.1597466961874381e-05,
+      "loss": 0.5131,
+      "step": 6000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2082339717307802,
+      "learning_rate": 1.1594986508015698e-05,
+      "loss": 0.555,
+      "step": 6001
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1382908911679626,
+      "learning_rate": 1.1592505953457282e-05,
+      "loss": 0.6189,
+      "step": 6002
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3104182491484984,
+      "learning_rate": 1.1590025298355749e-05,
+      "loss": 0.5667,
+      "step": 6003
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3662180924245277,
+      "learning_rate": 1.1587544542867716e-05,
+      "loss": 0.6686,
+      "step": 6004
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.151598103980875,
+      "learning_rate": 1.1585063687149807e-05,
+      "loss": 0.5373,
+      "step": 6005
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1814977133846232,
+      "learning_rate": 1.158258273135865e-05,
+      "loss": 0.5538,
+      "step": 6006
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.224022391516502,
+      "learning_rate": 1.158010167565088e-05,
+      "loss": 0.6384,
+      "step": 6007
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1989991424977446,
+      "learning_rate": 1.1577620520183135e-05,
+      "loss": 0.5239,
+      "step": 6008
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1996822393430697,
+      "learning_rate": 1.1575139265112072e-05,
+      "loss": 0.5065,
+      "step": 6009
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1306896444705934,
+      "learning_rate": 1.1572657910594336e-05,
+      "loss": 0.5249,
+      "step": 6010
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1442817702513723,
+      "learning_rate": 1.1570176456786597e-05,
+      "loss": 0.5541,
+      "step": 6011
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2164129199443567,
+      "learning_rate": 1.1567694903845515e-05,
+      "loss": 0.5739,
+      "step": 6012
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2389645781914767,
+      "learning_rate": 1.1565213251927764e-05,
+      "loss": 0.5953,
+      "step": 6013
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3176629381565335,
+      "learning_rate": 1.1562731501190027e-05,
+      "loss": 0.6181,
+      "step": 6014
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.080050587087489,
+      "learning_rate": 1.1560249651788985e-05,
+      "loss": 0.4979,
+      "step": 6015
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.213336610728508,
+      "learning_rate": 1.155776770388134e-05,
+      "loss": 0.5931,
+      "step": 6016
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2257029929626926,
+      "learning_rate": 1.1555285657623776e-05,
+      "loss": 0.5298,
+      "step": 6017
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.199148766280429,
+      "learning_rate": 1.1552803513173004e-05,
+      "loss": 0.5782,
+      "step": 6018
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1464487099502134,
+      "learning_rate": 1.1550321270685739e-05,
+      "loss": 0.5308,
+      "step": 6019
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2900742230760318,
+      "learning_rate": 1.1547838930318689e-05,
+      "loss": 0.5616,
+      "step": 6020
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.181885143863875,
+      "learning_rate": 1.1545356492228585e-05,
+      "loss": 0.5739,
+      "step": 6021
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1470877052572075,
+      "learning_rate": 1.1542873956572151e-05,
+      "loss": 0.5653,
+      "step": 6022
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0975241802219553,
+      "learning_rate": 1.1540391323506124e-05,
+      "loss": 0.5422,
+      "step": 6023
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2671880606804111,
+      "learning_rate": 1.1537908593187246e-05,
+      "loss": 0.5618,
+      "step": 6024
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2110152188863899,
+      "learning_rate": 1.1535425765772262e-05,
+      "loss": 0.5546,
+      "step": 6025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1214674482991491,
+      "learning_rate": 1.1532942841417931e-05,
+      "loss": 0.5415,
+      "step": 6026
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3119752379819598,
+      "learning_rate": 1.1530459820281008e-05,
+      "loss": 0.6052,
+      "step": 6027
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1237423542853127,
+      "learning_rate": 1.1527976702518257e-05,
+      "loss": 0.5355,
+      "step": 6028
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1256897189583506,
+      "learning_rate": 1.1525493488286458e-05,
+      "loss": 0.5348,
+      "step": 6029
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2458725018335104,
+      "learning_rate": 1.1523010177742382e-05,
+      "loss": 0.565,
+      "step": 6030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1306324996404287,
+      "learning_rate": 1.152052677104282e-05,
+      "loss": 0.5562,
+      "step": 6031
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1892979212134052,
+      "learning_rate": 1.1518043268344554e-05,
+      "loss": 0.6081,
+      "step": 6032
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2624901459564288,
+      "learning_rate": 1.1515559669804386e-05,
+      "loss": 0.6001,
+      "step": 6033
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1884519876965736,
+      "learning_rate": 1.1513075975579116e-05,
+      "loss": 0.5533,
+      "step": 6034
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0712219606733848,
+      "learning_rate": 1.1510592185825553e-05,
+      "loss": 0.4657,
+      "step": 6035
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.229893674131583,
+      "learning_rate": 1.150810830070051e-05,
+      "loss": 0.5969,
+      "step": 6036
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1520841144377467,
+      "learning_rate": 1.1505624320360813e-05,
+      "loss": 0.5821,
+      "step": 6037
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1657764694346948,
+      "learning_rate": 1.1503140244963283e-05,
+      "loss": 0.5733,
+      "step": 6038
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1349233494726247,
+      "learning_rate": 1.1500656074664756e-05,
+      "loss": 0.5414,
+      "step": 6039
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1861377732920508,
+      "learning_rate": 1.1498171809622067e-05,
+      "loss": 0.5422,
+      "step": 6040
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.150809763796626,
+      "learning_rate": 1.1495687449992059e-05,
+      "loss": 0.4986,
+      "step": 6041
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1371039298375025,
+      "learning_rate": 1.149320299593159e-05,
+      "loss": 0.5916,
+      "step": 6042
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1367008692852423,
+      "learning_rate": 1.149071844759751e-05,
+      "loss": 0.5277,
+      "step": 6043
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1330809702580076,
+      "learning_rate": 1.1488233805146685e-05,
+      "loss": 0.5555,
+      "step": 6044
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.186683273538549,
+      "learning_rate": 1.1485749068735982e-05,
+      "loss": 0.5772,
+      "step": 6045
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1466717802440118,
+      "learning_rate": 1.148326423852227e-05,
+      "loss": 0.5352,
+      "step": 6046
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.167547807143518,
+      "learning_rate": 1.1480779314662438e-05,
+      "loss": 0.5804,
+      "step": 6047
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.290652733382439,
+      "learning_rate": 1.1478294297313366e-05,
+      "loss": 0.6415,
+      "step": 6048
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1767032088278035,
+      "learning_rate": 1.1475809186631947e-05,
+      "loss": 0.6049,
+      "step": 6049
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.129735094600574,
+      "learning_rate": 1.1473323982775085e-05,
+      "loss": 0.5539,
+      "step": 6050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.177364715207251,
+      "learning_rate": 1.1470838685899675e-05,
+      "loss": 0.532,
+      "step": 6051
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1061015121030253,
+      "learning_rate": 1.146835329616263e-05,
+      "loss": 0.5021,
+      "step": 6052
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1822799592411755,
+      "learning_rate": 1.1465867813720865e-05,
+      "loss": 0.5967,
+      "step": 6053
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1386216743094857,
+      "learning_rate": 1.1463382238731305e-05,
+      "loss": 0.5561,
+      "step": 6054
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1723453849243168,
+      "learning_rate": 1.1460896571350875e-05,
+      "loss": 0.5998,
+      "step": 6055
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.210706331891644,
+      "learning_rate": 1.1458410811736503e-05,
+      "loss": 0.5615,
+      "step": 6056
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1605290286021857,
+      "learning_rate": 1.1455924960045136e-05,
+      "loss": 0.5542,
+      "step": 6057
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3127881142599516,
+      "learning_rate": 1.145343901643371e-05,
+      "loss": 0.5599,
+      "step": 6058
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1901311836542792,
+      "learning_rate": 1.1450952981059182e-05,
+      "loss": 0.5495,
+      "step": 6059
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2635045124995357,
+      "learning_rate": 1.144846685407851e-05,
+      "loss": 0.623,
+      "step": 6060
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.236935868425668,
+      "learning_rate": 1.1445980635648649e-05,
+      "loss": 0.5437,
+      "step": 6061
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1302637827876,
+      "learning_rate": 1.1443494325926572e-05,
+      "loss": 0.5638,
+      "step": 6062
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2087317774443476,
+      "learning_rate": 1.1441007925069248e-05,
+      "loss": 0.5703,
+      "step": 6063
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2696397470915577,
+      "learning_rate": 1.1438521433233657e-05,
+      "loss": 0.5883,
+      "step": 6064
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0951124697019443,
+      "learning_rate": 1.1436034850576794e-05,
+      "loss": 0.5382,
+      "step": 6065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0863121539069758,
+      "learning_rate": 1.1433548177255638e-05,
+      "loss": 0.5102,
+      "step": 6066
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2123883815396206,
+      "learning_rate": 1.143106141342719e-05,
+      "loss": 0.5828,
+      "step": 6067
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1480736707615016,
+      "learning_rate": 1.1428574559248448e-05,
+      "loss": 0.5817,
+      "step": 6068
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1037466701919598,
+      "learning_rate": 1.1426087614876424e-05,
+      "loss": 0.5306,
+      "step": 6069
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.171892750923502,
+      "learning_rate": 1.1423600580468137e-05,
+      "loss": 0.5238,
+      "step": 6070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.21161450565429,
+      "learning_rate": 1.1421113456180597e-05,
+      "loss": 0.5793,
+      "step": 6071
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.145258909795778,
+      "learning_rate": 1.1418626242170833e-05,
+      "loss": 0.5717,
+      "step": 6072
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1455312793145906,
+      "learning_rate": 1.1416138938595874e-05,
+      "loss": 0.5468,
+      "step": 6073
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1801996193160358,
+      "learning_rate": 1.1413651545612758e-05,
+      "loss": 0.566,
+      "step": 6074
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1543277536254237,
+      "learning_rate": 1.1411164063378529e-05,
+      "loss": 0.5614,
+      "step": 6075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2549773303108371,
+      "learning_rate": 1.1408676492050229e-05,
+      "loss": 0.5937,
+      "step": 6076
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0943050610967415,
+      "learning_rate": 1.1406188831784912e-05,
+      "loss": 0.5143,
+      "step": 6077
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2521838185404046,
+      "learning_rate": 1.1403701082739644e-05,
+      "loss": 0.6241,
+      "step": 6078
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2241330231799306,
+      "learning_rate": 1.1401213245071481e-05,
+      "loss": 0.591,
+      "step": 6079
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1258022838483297,
+      "learning_rate": 1.1398725318937503e-05,
+      "loss": 0.5129,
+      "step": 6080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.251965408145916,
+      "learning_rate": 1.1396237304494772e-05,
+      "loss": 0.5122,
+      "step": 6081
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1887100980782717,
+      "learning_rate": 1.1393749201900377e-05,
+      "loss": 0.5747,
+      "step": 6082
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1874656672282982,
+      "learning_rate": 1.1391261011311407e-05,
+      "loss": 0.513,
+      "step": 6083
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0929043361614448,
+      "learning_rate": 1.138877273288495e-05,
+      "loss": 0.5162,
+      "step": 6084
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0594942549027224,
+      "learning_rate": 1.1386284366778106e-05,
+      "loss": 0.506,
+      "step": 6085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.237677345669188,
+      "learning_rate": 1.1383795913147978e-05,
+      "loss": 0.5822,
+      "step": 6086
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0246579413587495,
+      "learning_rate": 1.138130737215167e-05,
+      "loss": 0.4975,
+      "step": 6087
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1503271010885996,
+      "learning_rate": 1.1378818743946308e-05,
+      "loss": 0.5731,
+      "step": 6088
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3168991931780432,
+      "learning_rate": 1.1376330028689e-05,
+      "loss": 0.54,
+      "step": 6089
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1430622346810895,
+      "learning_rate": 1.137384122653688e-05,
+      "loss": 0.516,
+      "step": 6090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0391458463584249,
+      "learning_rate": 1.137135233764707e-05,
+      "loss": 0.5263,
+      "step": 6091
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2606152412214606,
+      "learning_rate": 1.1368863362176713e-05,
+      "loss": 0.5872,
+      "step": 6092
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2150401848948964,
+      "learning_rate": 1.1366374300282954e-05,
+      "loss": 0.5402,
+      "step": 6093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2435444551780566,
+      "learning_rate": 1.1363885152122933e-05,
+      "loss": 0.5731,
+      "step": 6094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1744543553654667,
+      "learning_rate": 1.1361395917853808e-05,
+      "loss": 0.5379,
+      "step": 6095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2422188329043826,
+      "learning_rate": 1.1358906597632731e-05,
+      "loss": 0.5761,
+      "step": 6096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.169244535132021,
+      "learning_rate": 1.135641719161687e-05,
+      "loss": 0.5983,
+      "step": 6097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1300913256478453,
+      "learning_rate": 1.1353927699963396e-05,
+      "loss": 0.5251,
+      "step": 6098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2876857456920567,
+      "learning_rate": 1.135143812282948e-05,
+      "loss": 0.5808,
+      "step": 6099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1864677761507028,
+      "learning_rate": 1.1348948460372302e-05,
+      "loss": 0.5515,
+      "step": 6100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0960707430284047,
+      "learning_rate": 1.1346458712749049e-05,
+      "loss": 0.5575,
+      "step": 6101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1661438735643315,
+      "learning_rate": 1.1343968880116907e-05,
+      "loss": 0.5612,
+      "step": 6102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1487139414543932,
+      "learning_rate": 1.1341478962633081e-05,
+      "loss": 0.5233,
+      "step": 6103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0363604756585434,
+      "learning_rate": 1.1338988960454763e-05,
+      "loss": 0.4851,
+      "step": 6104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1794222603017963,
+      "learning_rate": 1.1336498873739166e-05,
+      "loss": 0.5996,
+      "step": 6105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2842968789355962,
+      "learning_rate": 1.13340087026435e-05,
+      "loss": 0.6117,
+      "step": 6106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1565401383397298,
+      "learning_rate": 1.1331518447324978e-05,
+      "loss": 0.5492,
+      "step": 6107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2901046239579763,
+      "learning_rate": 1.1329028107940832e-05,
+      "loss": 0.6147,
+      "step": 6108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0247321057884973,
+      "learning_rate": 1.1326537684648282e-05,
+      "loss": 0.5042,
+      "step": 6109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1642414441915516,
+      "learning_rate": 1.1324047177604565e-05,
+      "loss": 0.5156,
+      "step": 6110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2482839726279775,
+      "learning_rate": 1.1321556586966917e-05,
+      "loss": 0.5936,
+      "step": 6111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1575613060817325,
+      "learning_rate": 1.1319065912892584e-05,
+      "loss": 0.5384,
+      "step": 6112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2922929455541055,
+      "learning_rate": 1.1316575155538816e-05,
+      "loss": 0.6094,
+      "step": 6113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1511815821792062,
+      "learning_rate": 1.1314084315062863e-05,
+      "loss": 0.5347,
+      "step": 6114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1563639713512337,
+      "learning_rate": 1.131159339162199e-05,
+      "loss": 0.5689,
+      "step": 6115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1636079758498936,
+      "learning_rate": 1.1309102385373459e-05,
+      "loss": 0.5629,
+      "step": 6116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2872494518599655,
+      "learning_rate": 1.1306611296474536e-05,
+      "loss": 0.5658,
+      "step": 6117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1639668982246267,
+      "learning_rate": 1.1304120125082504e-05,
+      "loss": 0.5931,
+      "step": 6118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1237137116714113,
+      "learning_rate": 1.1301628871354641e-05,
+      "loss": 0.5275,
+      "step": 6119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1171285973873313,
+      "learning_rate": 1.129913753544823e-05,
+      "loss": 0.5402,
+      "step": 6120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2248233687287509,
+      "learning_rate": 1.1296646117520567e-05,
+      "loss": 0.5971,
+      "step": 6121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.144457557571205,
+      "learning_rate": 1.1294154617728942e-05,
+      "loss": 0.5616,
+      "step": 6122
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.128843206737472,
+      "learning_rate": 1.1291663036230658e-05,
+      "loss": 0.6098,
+      "step": 6123
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1859439392317106,
+      "learning_rate": 1.1289171373183026e-05,
+      "loss": 0.5655,
+      "step": 6124
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2119237021801355,
+      "learning_rate": 1.1286679628743349e-05,
+      "loss": 0.5533,
+      "step": 6125
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2164796565746279,
+      "learning_rate": 1.1284187803068953e-05,
+      "loss": 0.6083,
+      "step": 6126
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1327797194374818,
+      "learning_rate": 1.1281695896317153e-05,
+      "loss": 0.5423,
+      "step": 6127
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2823281287869477,
+      "learning_rate": 1.127920390864528e-05,
+      "loss": 0.5764,
+      "step": 6128
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.070036748183764,
+      "learning_rate": 1.1276711840210663e-05,
+      "loss": 0.5204,
+      "step": 6129
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2320117781992606,
+      "learning_rate": 1.127421969117064e-05,
+      "loss": 0.5419,
+      "step": 6130
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1750960270215633,
+      "learning_rate": 1.1271727461682558e-05,
+      "loss": 0.5746,
+      "step": 6131
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2556764935216203,
+      "learning_rate": 1.1269235151903754e-05,
+      "loss": 0.5761,
+      "step": 6132
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1858653814328417,
+      "learning_rate": 1.126674276199159e-05,
+      "loss": 0.5399,
+      "step": 6133
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.120311610435654,
+      "learning_rate": 1.1264250292103423e-05,
+      "loss": 0.4987,
+      "step": 6134
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2430094272903576,
+      "learning_rate": 1.1261757742396606e-05,
+      "loss": 0.6163,
+      "step": 6135
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1713331877357887,
+      "learning_rate": 1.1259265113028517e-05,
+      "loss": 0.567,
+      "step": 6136
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2023741125247693,
+      "learning_rate": 1.1256772404156521e-05,
+      "loss": 0.5412,
+      "step": 6137
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1715327462452352,
+      "learning_rate": 1.1254279615938001e-05,
+      "loss": 0.603,
+      "step": 6138
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1307029813630516,
+      "learning_rate": 1.1251786748530342e-05,
+      "loss": 0.5531,
+      "step": 6139
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1911774962897999,
+      "learning_rate": 1.124929380209092e-05,
+      "loss": 0.5529,
+      "step": 6140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1671098593916343,
+      "learning_rate": 1.124680077677714e-05,
+      "loss": 0.5478,
+      "step": 6141
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.209040528011753,
+      "learning_rate": 1.124430767274639e-05,
+      "loss": 0.6081,
+      "step": 6142
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1465392221462014,
+      "learning_rate": 1.124181449015608e-05,
+      "loss": 0.58,
+      "step": 6143
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.131175627819122,
+      "learning_rate": 1.1239321229163615e-05,
+      "loss": 0.5669,
+      "step": 6144
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2186527213374134,
+      "learning_rate": 1.1236827889926402e-05,
+      "loss": 0.5652,
+      "step": 6145
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2058958867967815,
+      "learning_rate": 1.1234334472601868e-05,
+      "loss": 0.607,
+      "step": 6146
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2508591560343088,
+      "learning_rate": 1.1231840977347427e-05,
+      "loss": 0.5752,
+      "step": 6147
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3113149561034705,
+      "learning_rate": 1.1229347404320515e-05,
+      "loss": 0.6231,
+      "step": 6148
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1797706119994518,
+      "learning_rate": 1.1226853753678555e-05,
+      "loss": 0.5521,
+      "step": 6149
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1573637675782535,
+      "learning_rate": 1.1224360025578987e-05,
+      "loss": 0.5661,
+      "step": 6150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0177478612110338,
+      "learning_rate": 1.1221866220179254e-05,
+      "loss": 0.5118,
+      "step": 6151
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2031172714976002,
+      "learning_rate": 1.1219372337636802e-05,
+      "loss": 0.5805,
+      "step": 6152
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1654660768551133,
+      "learning_rate": 1.1216878378109085e-05,
+      "loss": 0.591,
+      "step": 6153
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.278893000791008,
+      "learning_rate": 1.1214384341753557e-05,
+      "loss": 0.5893,
+      "step": 6154
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0944814688384625,
+      "learning_rate": 1.1211890228727679e-05,
+      "loss": 0.5593,
+      "step": 6155
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2107200673460006,
+      "learning_rate": 1.120939603918892e-05,
+      "loss": 0.5401,
+      "step": 6156
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1891111186009367,
+      "learning_rate": 1.1206901773294749e-05,
+      "loss": 0.5467,
+      "step": 6157
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2391264524948702,
+      "learning_rate": 1.1204407431202642e-05,
+      "loss": 0.5943,
+      "step": 6158
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1448040009037483,
+      "learning_rate": 1.120191301307008e-05,
+      "loss": 0.553,
+      "step": 6159
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0451598279734644,
+      "learning_rate": 1.1199418519054549e-05,
+      "loss": 0.4847,
+      "step": 6160
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1327962939977638,
+      "learning_rate": 1.1196923949313537e-05,
+      "loss": 0.5197,
+      "step": 6161
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1831119173498421,
+      "learning_rate": 1.1194429304004541e-05,
+      "loss": 0.5999,
+      "step": 6162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.211938997637693,
+      "learning_rate": 1.1191934583285063e-05,
+      "loss": 0.5342,
+      "step": 6163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0937776834526711,
+      "learning_rate": 1.1189439787312603e-05,
+      "loss": 0.5601,
+      "step": 6164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1182546787108651,
+      "learning_rate": 1.1186944916244673e-05,
+      "loss": 0.5559,
+      "step": 6165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4341959922639842,
+      "learning_rate": 1.1184449970238787e-05,
+      "loss": 0.5554,
+      "step": 6166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1721113856638508,
+      "learning_rate": 1.1181954949452463e-05,
+      "loss": 0.5801,
+      "step": 6167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1589267265091858,
+      "learning_rate": 1.1179459854043227e-05,
+      "loss": 0.5391,
+      "step": 6168
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1216112597206436,
+      "learning_rate": 1.1176964684168603e-05,
+      "loss": 0.5426,
+      "step": 6169
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1056470962998983,
+      "learning_rate": 1.1174469439986126e-05,
+      "loss": 0.5618,
+      "step": 6170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2247803004468978,
+      "learning_rate": 1.1171974121653333e-05,
+      "loss": 0.5663,
+      "step": 6171
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1264885485287592,
+      "learning_rate": 1.116947872932777e-05,
+      "loss": 0.5633,
+      "step": 6172
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1898257919829878,
+      "learning_rate": 1.1166983263166979e-05,
+      "loss": 0.5433,
+      "step": 6173
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1917158381225335,
+      "learning_rate": 1.1164487723328516e-05,
+      "loss": 0.5849,
+      "step": 6174
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0908833675891934,
+      "learning_rate": 1.1161992109969932e-05,
+      "loss": 0.5728,
+      "step": 6175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2274191228012494,
+      "learning_rate": 1.115949642324879e-05,
+      "loss": 0.6096,
+      "step": 6176
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.140941001282802,
+      "learning_rate": 1.1157000663322662e-05,
+      "loss": 0.5523,
+      "step": 6177
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2186166250168242,
+      "learning_rate": 1.115450483034911e-05,
+      "loss": 0.5749,
+      "step": 6178
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1872916289770912,
+      "learning_rate": 1.115200892448571e-05,
+      "loss": 0.5628,
+      "step": 6179
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1816432479955392,
+      "learning_rate": 1.1149512945890044e-05,
+      "loss": 0.5599,
+      "step": 6180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1784970674959776,
+      "learning_rate": 1.1147016894719695e-05,
+      "loss": 0.5754,
+      "step": 6181
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1310563253615806,
+      "learning_rate": 1.1144520771132252e-05,
+      "loss": 0.534,
+      "step": 6182
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1835764767233026,
+      "learning_rate": 1.1142024575285308e-05,
+      "loss": 0.5458,
+      "step": 6183
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1392449168798588,
+      "learning_rate": 1.1139528307336463e-05,
+      "loss": 0.5747,
+      "step": 6184
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.158187634523703,
+      "learning_rate": 1.1137031967443312e-05,
+      "loss": 0.5877,
+      "step": 6185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2891116046222504,
+      "learning_rate": 1.1134535555763466e-05,
+      "loss": 0.5957,
+      "step": 6186
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1423768999322064,
+      "learning_rate": 1.113203907245454e-05,
+      "loss": 0.5516,
+      "step": 6187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1826018152096238,
+      "learning_rate": 1.1129542517674147e-05,
+      "loss": 0.57,
+      "step": 6188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1678530020117521,
+      "learning_rate": 1.1127045891579906e-05,
+      "loss": 0.4987,
+      "step": 6189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1482155514031898,
+      "learning_rate": 1.1124549194329445e-05,
+      "loss": 0.5555,
+      "step": 6190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1638159778875492,
+      "learning_rate": 1.112205242608039e-05,
+      "loss": 0.5227,
+      "step": 6191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3314113572883042,
+      "learning_rate": 1.1119555586990376e-05,
+      "loss": 0.6453,
+      "step": 6192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1504087592413548,
+      "learning_rate": 1.1117058677217043e-05,
+      "loss": 0.5506,
+      "step": 6193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1109482261687162,
+      "learning_rate": 1.111456169691803e-05,
+      "loss": 0.5282,
+      "step": 6194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2655304767491267,
+      "learning_rate": 1.1112064646250988e-05,
+      "loss": 0.5556,
+      "step": 6195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1314574457140631,
+      "learning_rate": 1.1109567525373569e-05,
+      "loss": 0.4926,
+      "step": 6196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2467107889370879,
+      "learning_rate": 1.1107070334443426e-05,
+      "loss": 0.6065,
+      "step": 6197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.098095083247121,
+      "learning_rate": 1.1104573073618222e-05,
+      "loss": 0.5326,
+      "step": 6198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0949034467133698,
+      "learning_rate": 1.1102075743055618e-05,
+      "loss": 0.5128,
+      "step": 6199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2051167536005194,
+      "learning_rate": 1.1099578342913289e-05,
+      "loss": 0.5695,
+      "step": 6200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1850271075403587,
+      "learning_rate": 1.1097080873348905e-05,
+      "loss": 0.598,
+      "step": 6201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.270252574525389,
+      "learning_rate": 1.1094583334520146e-05,
+      "loss": 0.6388,
+      "step": 6202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1029827620576405,
+      "learning_rate": 1.1092085726584693e-05,
+      "loss": 0.5326,
+      "step": 6203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1215251664064685,
+      "learning_rate": 1.1089588049700234e-05,
+      "loss": 0.5437,
+      "step": 6204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.155188459816482,
+      "learning_rate": 1.1087090304024462e-05,
+      "loss": 0.544,
+      "step": 6205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1474209912428457,
+      "learning_rate": 1.1084592489715067e-05,
+      "loss": 0.5899,
+      "step": 6206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2369202556301464,
+      "learning_rate": 1.1082094606929754e-05,
+      "loss": 0.5627,
+      "step": 6207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2608004792179852,
+      "learning_rate": 1.1079596655826227e-05,
+      "loss": 0.5612,
+      "step": 6208
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1596070604892992,
+      "learning_rate": 1.1077098636562191e-05,
+      "loss": 0.5742,
+      "step": 6209
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1999462791180133,
+      "learning_rate": 1.1074600549295363e-05,
+      "loss": 0.5375,
+      "step": 6210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1406132945675063,
+      "learning_rate": 1.1072102394183456e-05,
+      "loss": 0.6112,
+      "step": 6211
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1389317420454288,
+      "learning_rate": 1.1069604171384194e-05,
+      "loss": 0.5766,
+      "step": 6212
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.21567475087168,
+      "learning_rate": 1.1067105881055303e-05,
+      "loss": 0.61,
+      "step": 6213
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1913558198855603,
+      "learning_rate": 1.106460752335451e-05,
+      "loss": 0.5465,
+      "step": 6214
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2962121763385255,
+      "learning_rate": 1.1062109098439555e-05,
+      "loss": 0.5838,
+      "step": 6215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1993759996530284,
+      "learning_rate": 1.105961060646817e-05,
+      "loss": 0.5563,
+      "step": 6216
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2047749884634675,
+      "learning_rate": 1.1057112047598102e-05,
+      "loss": 0.5677,
+      "step": 6217
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0570390271043815,
+      "learning_rate": 1.1054613421987098e-05,
+      "loss": 0.5383,
+      "step": 6218
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.188914761281565,
+      "learning_rate": 1.1052114729792902e-05,
+      "loss": 0.5815,
+      "step": 6219
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2046005811394422,
+      "learning_rate": 1.104961597117328e-05,
+      "loss": 0.5691,
+      "step": 6220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1500618275317596,
+      "learning_rate": 1.1047117146285984e-05,
+      "loss": 0.5111,
+      "step": 6221
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0534887212105184,
+      "learning_rate": 1.1044618255288781e-05,
+      "loss": 0.5623,
+      "step": 6222
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1762681124161223,
+      "learning_rate": 1.104211929833944e-05,
+      "loss": 0.54,
+      "step": 6223
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0849735729673213,
+      "learning_rate": 1.1039620275595729e-05,
+      "loss": 0.5644,
+      "step": 6224
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.082915182598213,
+      "learning_rate": 1.1037121187215427e-05,
+      "loss": 0.5286,
+      "step": 6225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1481968114964451,
+      "learning_rate": 1.1034622033356311e-05,
+      "loss": 0.5686,
+      "step": 6226
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1205564799851595,
+      "learning_rate": 1.103212281417617e-05,
+      "loss": 0.5383,
+      "step": 6227
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0921265133487799,
+      "learning_rate": 1.1029623529832793e-05,
+      "loss": 0.5392,
+      "step": 6228
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.04841506183565,
+      "learning_rate": 1.1027124180483965e-05,
+      "loss": 0.5416,
+      "step": 6229
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.168596029084298,
+      "learning_rate": 1.1024624766287492e-05,
+      "loss": 0.5397,
+      "step": 6230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1717812055563395,
+      "learning_rate": 1.1022125287401172e-05,
+      "loss": 0.5511,
+      "step": 6231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.248813256541239,
+      "learning_rate": 1.1019625743982807e-05,
+      "loss": 0.5685,
+      "step": 6232
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1908274761257898,
+      "learning_rate": 1.101712613619021e-05,
+      "loss": 0.5442,
+      "step": 6233
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1641961347436167,
+      "learning_rate": 1.1014626464181191e-05,
+      "loss": 0.543,
+      "step": 6234
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.084253556104743,
+      "learning_rate": 1.1012126728113567e-05,
+      "loss": 0.4856,
+      "step": 6235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1528492998561048,
+      "learning_rate": 1.1009626928145163e-05,
+      "loss": 0.6098,
+      "step": 6236
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.146757700895703,
+      "learning_rate": 1.1007127064433802e-05,
+      "loss": 0.5844,
+      "step": 6237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1120951440916955,
+      "learning_rate": 1.1004627137137314e-05,
+      "loss": 0.5116,
+      "step": 6238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1138632340672925,
+      "learning_rate": 1.1002127146413531e-05,
+      "loss": 0.5143,
+      "step": 6239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2878599163787527,
+      "learning_rate": 1.0999627092420291e-05,
+      "loss": 0.5805,
+      "step": 6240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2106485329675938,
+      "learning_rate": 1.0997126975315433e-05,
+      "loss": 0.5232,
+      "step": 6241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0727876320192133,
+      "learning_rate": 1.0994626795256806e-05,
+      "loss": 0.5055,
+      "step": 6242
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1713025538685444,
+      "learning_rate": 1.0992126552402261e-05,
+      "loss": 0.5475,
+      "step": 6243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.263902646060917,
+      "learning_rate": 1.0989626246909642e-05,
+      "loss": 0.5828,
+      "step": 6244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2439657474659542,
+      "learning_rate": 1.0987125878936814e-05,
+      "loss": 0.5851,
+      "step": 6245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.193978556378137,
+      "learning_rate": 1.0984625448641639e-05,
+      "loss": 0.5448,
+      "step": 6246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.191254002583395,
+      "learning_rate": 1.0982124956181979e-05,
+      "loss": 0.5728,
+      "step": 6247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1800802726435635,
+      "learning_rate": 1.0979624401715702e-05,
+      "loss": 0.5526,
+      "step": 6248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.193635449730372,
+      "learning_rate": 1.0977123785400684e-05,
+      "loss": 0.5814,
+      "step": 6249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2722000006762226,
+      "learning_rate": 1.0974623107394797e-05,
+      "loss": 0.5663,
+      "step": 6250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2157040705130804,
+      "learning_rate": 1.097212236785593e-05,
+      "loss": 0.5649,
+      "step": 6251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.282610519807765,
+      "learning_rate": 1.0969621566941959e-05,
+      "loss": 0.5823,
+      "step": 6252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0865906322437042,
+      "learning_rate": 1.0967120704810776e-05,
+      "loss": 0.5579,
+      "step": 6253
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1548739828581531,
+      "learning_rate": 1.0964619781620271e-05,
+      "loss": 0.5986,
+      "step": 6254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2150745233985072,
+      "learning_rate": 1.0962118797528344e-05,
+      "loss": 0.5374,
+      "step": 6255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.15572927966926,
+      "learning_rate": 1.0959617752692897e-05,
+      "loss": 0.5211,
+      "step": 6256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1692748151115242,
+      "learning_rate": 1.0957116647271829e-05,
+      "loss": 0.5543,
+      "step": 6257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0414781527011072,
+      "learning_rate": 1.0954615481423047e-05,
+      "loss": 0.5168,
+      "step": 6258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1498510782594316,
+      "learning_rate": 1.0952114255304465e-05,
+      "loss": 0.5256,
+      "step": 6259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1825853338763397,
+      "learning_rate": 1.0949612969073995e-05,
+      "loss": 0.5309,
+      "step": 6260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2336561789971732,
+      "learning_rate": 1.0947111622889563e-05,
+      "loss": 0.6284,
+      "step": 6261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1557986950972743,
+      "learning_rate": 1.0944610216909086e-05,
+      "loss": 0.5068,
+      "step": 6262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.30853617527261,
+      "learning_rate": 1.0942108751290494e-05,
+      "loss": 0.599,
+      "step": 6263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.068262461951388,
+      "learning_rate": 1.0939607226191716e-05,
+      "loss": 0.5627,
+      "step": 6264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1874667213193926,
+      "learning_rate": 1.0937105641770682e-05,
+      "loss": 0.5241,
+      "step": 6265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1902940405277826,
+      "learning_rate": 1.0934603998185338e-05,
+      "loss": 0.5698,
+      "step": 6266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0901840936430078,
+      "learning_rate": 1.0932102295593621e-05,
+      "loss": 0.5125,
+      "step": 6267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1855604494628034,
+      "learning_rate": 1.0929600534153477e-05,
+      "loss": 0.5043,
+      "step": 6268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1662671506383175,
+      "learning_rate": 1.0927098714022854e-05,
+      "loss": 0.5761,
+      "step": 6269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1207815659396392,
+      "learning_rate": 1.0924596835359706e-05,
+      "loss": 0.5716,
+      "step": 6270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0947266713941182,
+      "learning_rate": 1.092209489832199e-05,
+      "loss": 0.5675,
+      "step": 6271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.327508087492767,
+      "learning_rate": 1.0919592903067668e-05,
+      "loss": 0.6027,
+      "step": 6272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1924033073394782,
+      "learning_rate": 1.09170908497547e-05,
+      "loss": 0.5869,
+      "step": 6273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0611016664755974,
+      "learning_rate": 1.0914588738541055e-05,
+      "loss": 0.5745,
+      "step": 6274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0868997495793626,
+      "learning_rate": 1.0912086569584702e-05,
+      "loss": 0.537,
+      "step": 6275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1296068281438767,
+      "learning_rate": 1.090958434304362e-05,
+      "loss": 0.5627,
+      "step": 6276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2191840645953926,
+      "learning_rate": 1.0907082059075786e-05,
+      "loss": 0.5763,
+      "step": 6277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.052527922307498,
+      "learning_rate": 1.090457971783918e-05,
+      "loss": 0.53,
+      "step": 6278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2259940998937324,
+      "learning_rate": 1.0902077319491792e-05,
+      "loss": 0.5499,
+      "step": 6279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2058268837711255,
+      "learning_rate": 1.0899574864191607e-05,
+      "loss": 0.5682,
+      "step": 6280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2088781750798747,
+      "learning_rate": 1.089707235209662e-05,
+      "loss": 0.603,
+      "step": 6281
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2780784337297917,
+      "learning_rate": 1.089456978336483e-05,
+      "loss": 0.6067,
+      "step": 6282
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.174701790810231,
+      "learning_rate": 1.089206715815423e-05,
+      "loss": 0.5304,
+      "step": 6283
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1634918452666847,
+      "learning_rate": 1.0889564476622829e-05,
+      "loss": 0.5746,
+      "step": 6284
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1119609834254902,
+      "learning_rate": 1.0887061738928632e-05,
+      "loss": 0.4633,
+      "step": 6285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1644985739083769,
+      "learning_rate": 1.0884558945229652e-05,
+      "loss": 0.5484,
+      "step": 6286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0371188003485314,
+      "learning_rate": 1.0882056095683905e-05,
+      "loss": 0.5112,
+      "step": 6287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0866193757464007,
+      "learning_rate": 1.0879553190449402e-05,
+      "loss": 0.5504,
+      "step": 6288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2967099923558516,
+      "learning_rate": 1.0877050229684169e-05,
+      "loss": 0.5938,
+      "step": 6289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1989126404982595,
+      "learning_rate": 1.087454721354623e-05,
+      "loss": 0.5652,
+      "step": 6290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1253676343619814,
+      "learning_rate": 1.0872044142193614e-05,
+      "loss": 0.5584,
+      "step": 6291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1026007987241102,
+      "learning_rate": 1.0869541015784353e-05,
+      "loss": 0.5484,
+      "step": 6292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0925516921150882,
+      "learning_rate": 1.0867037834476477e-05,
+      "loss": 0.5529,
+      "step": 6293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1207733228044954,
+      "learning_rate": 1.0864534598428034e-05,
+      "loss": 0.5793,
+      "step": 6294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2878947199211705,
+      "learning_rate": 1.0862031307797059e-05,
+      "loss": 0.5352,
+      "step": 6295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.162445891566968,
+      "learning_rate": 1.08595279627416e-05,
+      "loss": 0.5072,
+      "step": 6296
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2525412001086693,
+      "learning_rate": 1.085702456341971e-05,
+      "loss": 0.5884,
+      "step": 6297
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1914150550391847,
+      "learning_rate": 1.085452110998943e-05,
+      "loss": 0.5651,
+      "step": 6298
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.179156051066751,
+      "learning_rate": 1.0852017602608831e-05,
+      "loss": 0.5558,
+      "step": 6299
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1901141555110957,
+      "learning_rate": 1.084951404143596e-05,
+      "loss": 0.5582,
+      "step": 6300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1525455621414233,
+      "learning_rate": 1.0847010426628888e-05,
+      "loss": 0.5806,
+      "step": 6301
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.245375375345019,
+      "learning_rate": 1.0844506758345676e-05,
+      "loss": 0.5989,
+      "step": 6302
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2113059683073955,
+      "learning_rate": 1.0842003036744395e-05,
+      "loss": 0.605,
+      "step": 6303
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1326986321408101,
+      "learning_rate": 1.0839499261983119e-05,
+      "loss": 0.5538,
+      "step": 6304
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1612838762349662,
+      "learning_rate": 1.0836995434219924e-05,
+      "loss": 0.5823,
+      "step": 6305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0554097845970336,
+      "learning_rate": 1.0834491553612889e-05,
+      "loss": 0.5296,
+      "step": 6306
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.127822514185915,
+      "learning_rate": 1.0831987620320096e-05,
+      "loss": 0.5667,
+      "step": 6307
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2216793459762962,
+      "learning_rate": 1.0829483634499632e-05,
+      "loss": 0.6178,
+      "step": 6308
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0998509956356397,
+      "learning_rate": 1.0826979596309586e-05,
+      "loss": 0.5163,
+      "step": 6309
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2600095054479847,
+      "learning_rate": 1.082447550590805e-05,
+      "loss": 0.5673,
+      "step": 6310
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.112431626951256,
+      "learning_rate": 1.0821971363453125e-05,
+      "loss": 0.5487,
+      "step": 6311
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.133429418492436,
+      "learning_rate": 1.0819467169102906e-05,
+      "loss": 0.5242,
+      "step": 6312
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1067683799566865,
+      "learning_rate": 1.0816962923015495e-05,
+      "loss": 0.5654,
+      "step": 6313
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1814836382029623,
+      "learning_rate": 1.0814458625349002e-05,
+      "loss": 0.6146,
+      "step": 6314
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.197842482329181,
+      "learning_rate": 1.0811954276261532e-05,
+      "loss": 0.5133,
+      "step": 6315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2046544645655877,
+      "learning_rate": 1.08094498759112e-05,
+      "loss": 0.5578,
+      "step": 6316
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2002400078765783,
+      "learning_rate": 1.0806945424456124e-05,
+      "loss": 0.5531,
+      "step": 6317
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2547361770631573,
+      "learning_rate": 1.0804440922054415e-05,
+      "loss": 0.6205,
+      "step": 6318
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0131846526726394,
+      "learning_rate": 1.0801936368864203e-05,
+      "loss": 0.4707,
+      "step": 6319
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2048835782836915,
+      "learning_rate": 1.0799431765043609e-05,
+      "loss": 0.5665,
+      "step": 6320
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.12954234645605,
+      "learning_rate": 1.0796927110750766e-05,
+      "loss": 0.5378,
+      "step": 6321
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1835233963304486,
+      "learning_rate": 1.0794422406143802e-05,
+      "loss": 0.5567,
+      "step": 6322
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1189268750219665,
+      "learning_rate": 1.079191765138085e-05,
+      "loss": 0.5382,
+      "step": 6323
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.14656578697717,
+      "learning_rate": 1.0789412846620052e-05,
+      "loss": 0.561,
+      "step": 6324
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1947010944939058,
+      "learning_rate": 1.078690799201955e-05,
+      "loss": 0.5726,
+      "step": 6325
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1365450696890864,
+      "learning_rate": 1.0784403087737487e-05,
+      "loss": 0.5449,
+      "step": 6326
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1423639080370231,
+      "learning_rate": 1.078189813393201e-05,
+      "loss": 0.53,
+      "step": 6327
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0225251768785275,
+      "learning_rate": 1.0779393130761267e-05,
+      "loss": 0.5306,
+      "step": 6328
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.118557230362184,
+      "learning_rate": 1.0776888078383415e-05,
+      "loss": 0.51,
+      "step": 6329
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.182975532049757,
+      "learning_rate": 1.0774382976956613e-05,
+      "loss": 0.5317,
+      "step": 6330
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1382466433362692,
+      "learning_rate": 1.0771877826639016e-05,
+      "loss": 0.5587,
+      "step": 6331
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1963473558820212,
+      "learning_rate": 1.0769372627588792e-05,
+      "loss": 0.5636,
+      "step": 6332
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0906775235095851,
+      "learning_rate": 1.0766867379964101e-05,
+      "loss": 0.5241,
+      "step": 6333
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1583652736642014,
+      "learning_rate": 1.0764362083923117e-05,
+      "loss": 0.5167,
+      "step": 6334
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2013230143416829,
+      "learning_rate": 1.0761856739624012e-05,
+      "loss": 0.5582,
+      "step": 6335
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1879429743884447,
+      "learning_rate": 1.0759351347224961e-05,
+      "loss": 0.5507,
+      "step": 6336
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.153090102745418,
+      "learning_rate": 1.0756845906884141e-05,
+      "loss": 0.5871,
+      "step": 6337
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1621570737650624,
+      "learning_rate": 1.0754340418759734e-05,
+      "loss": 0.556,
+      "step": 6338
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1537395752077297,
+      "learning_rate": 1.0751834883009922e-05,
+      "loss": 0.5413,
+      "step": 6339
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2374101818126393,
+      "learning_rate": 1.0749329299792898e-05,
+      "loss": 0.6011,
+      "step": 6340
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1537678340471695,
+      "learning_rate": 1.074682366926685e-05,
+      "loss": 0.5396,
+      "step": 6341
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1881184724149834,
+      "learning_rate": 1.074431799158997e-05,
+      "loss": 0.5643,
+      "step": 6342
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1101484154808001,
+      "learning_rate": 1.0741812266920453e-05,
+      "loss": 0.5376,
+      "step": 6343
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2491237907270207,
+      "learning_rate": 1.0739306495416502e-05,
+      "loss": 0.578,
+      "step": 6344
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1629287011291416,
+      "learning_rate": 1.0736800677236316e-05,
+      "loss": 0.5562,
+      "step": 6345
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1693825728529563,
+      "learning_rate": 1.0734294812538105e-05,
+      "loss": 0.601,
+      "step": 6346
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.247201552688239,
+      "learning_rate": 1.0731788901480071e-05,
+      "loss": 0.5603,
+      "step": 6347
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1101095427339125,
+      "learning_rate": 1.072928294422043e-05,
+      "loss": 0.5114,
+      "step": 6348
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1596878596155982,
+      "learning_rate": 1.0726776940917391e-05,
+      "loss": 0.585,
+      "step": 6349
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1022584488792138,
+      "learning_rate": 1.0724270891729178e-05,
+      "loss": 0.5236,
+      "step": 6350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0470160204291132,
+      "learning_rate": 1.0721764796814005e-05,
+      "loss": 0.4791,
+      "step": 6351
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.113496297846422,
+      "learning_rate": 1.0719258656330095e-05,
+      "loss": 0.5885,
+      "step": 6352
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2683157868410164,
+      "learning_rate": 1.071675247043568e-05,
+      "loss": 0.6067,
+      "step": 6353
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1515656511323173,
+      "learning_rate": 1.0714246239288977e-05,
+      "loss": 0.5735,
+      "step": 6354
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.187368988037953,
+      "learning_rate": 1.0711739963048229e-05,
+      "loss": 0.5395,
+      "step": 6355
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1191194275983907,
+      "learning_rate": 1.0709233641871663e-05,
+      "loss": 0.567,
+      "step": 6356
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2201148484844067,
+      "learning_rate": 1.0706727275917519e-05,
+      "loss": 0.583,
+      "step": 6357
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1183576524556678,
+      "learning_rate": 1.0704220865344036e-05,
+      "loss": 0.5268,
+      "step": 6358
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1074173538680858,
+      "learning_rate": 1.0701714410309454e-05,
+      "loss": 0.5301,
+      "step": 6359
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2581336517718718,
+      "learning_rate": 1.0699207910972022e-05,
+      "loss": 0.6065,
+      "step": 6360
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.198044291898656,
+      "learning_rate": 1.069670136748999e-05,
+      "loss": 0.5631,
+      "step": 6361
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2029244887482289,
+      "learning_rate": 1.0694194780021603e-05,
+      "loss": 0.567,
+      "step": 6362
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1244606738611367,
+      "learning_rate": 1.069168814872512e-05,
+      "loss": 0.5642,
+      "step": 6363
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1542943964123136,
+      "learning_rate": 1.0689181473758793e-05,
+      "loss": 0.5415,
+      "step": 6364
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.106997830916901,
+      "learning_rate": 1.0686674755280886e-05,
+      "loss": 0.5411,
+      "step": 6365
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2113052794102166,
+      "learning_rate": 1.068416799344966e-05,
+      "loss": 0.5413,
+      "step": 6366
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.190706491560063,
+      "learning_rate": 1.0681661188423373e-05,
+      "loss": 0.5769,
+      "step": 6367
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1948511567253493,
+      "learning_rate": 1.0679154340360305e-05,
+      "loss": 0.5415,
+      "step": 6368
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2501847607441907,
+      "learning_rate": 1.0676647449418713e-05,
+      "loss": 0.5796,
+      "step": 6369
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1075798339141731,
+      "learning_rate": 1.067414051575688e-05,
+      "loss": 0.5408,
+      "step": 6370
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1636644745406046,
+      "learning_rate": 1.0671633539533082e-05,
+      "loss": 0.5405,
+      "step": 6371
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2107047072577046,
+      "learning_rate": 1.0669126520905588e-05,
+      "loss": 0.5397,
+      "step": 6372
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2012178737857797,
+      "learning_rate": 1.0666619460032688e-05,
+      "loss": 0.5712,
+      "step": 6373
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1286669942893288,
+      "learning_rate": 1.0664112357072658e-05,
+      "loss": 0.4973,
+      "step": 6374
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1936473842275321,
+      "learning_rate": 1.0661605212183791e-05,
+      "loss": 0.5289,
+      "step": 6375
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1477808729509356,
+      "learning_rate": 1.0659098025524374e-05,
+      "loss": 0.5972,
+      "step": 6376
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1864971142270393,
+      "learning_rate": 1.0656590797252697e-05,
+      "loss": 0.5544,
+      "step": 6377
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2993287921302266,
+      "learning_rate": 1.0654083527527056e-05,
+      "loss": 0.593,
+      "step": 6378
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2390407315010037,
+      "learning_rate": 1.0651576216505747e-05,
+      "loss": 0.5953,
+      "step": 6379
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2142161130993105,
+      "learning_rate": 1.0649068864347072e-05,
+      "loss": 0.5534,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2344725365069458,
+      "learning_rate": 1.0646561471209328e-05,
+      "loss": 0.5721,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2690621785922231,
+      "learning_rate": 1.0644054037250825e-05,
+      "loss": 0.577,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1976319790226606,
+      "learning_rate": 1.0641546562629865e-05,
+      "loss": 0.5779,
+      "step": 6383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.260962481056273,
+      "learning_rate": 1.0639039047504763e-05,
+      "loss": 0.5964,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1301798780290067,
+      "learning_rate": 1.0636531492033826e-05,
+      "loss": 0.5542,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2013236097315585,
+      "learning_rate": 1.0634023896375376e-05,
+      "loss": 0.5422,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.219123245208269,
+      "learning_rate": 1.0631516260687722e-05,
+      "loss": 0.592,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3344785967239943,
+      "learning_rate": 1.062900858512919e-05,
+      "loss": 0.5912,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.261559491100791,
+      "learning_rate": 1.06265008698581e-05,
+      "loss": 0.6231,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.128106120789023,
+      "learning_rate": 1.0623993115032781e-05,
+      "loss": 0.5315,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1808282250558813,
+      "learning_rate": 1.0621485320811552e-05,
+      "loss": 0.5405,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1439660665574554,
+      "learning_rate": 1.0618977487352752e-05,
+      "loss": 0.5359,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2414900063071381,
+      "learning_rate": 1.061646961481471e-05,
+      "loss": 0.6137,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.136545541682416,
+      "learning_rate": 1.0613961703355758e-05,
+      "loss": 0.5369,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.221299414908821,
+      "learning_rate": 1.0611453753134237e-05,
+      "loss": 0.6023,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0786747843665214,
+      "learning_rate": 1.060894576430849e-05,
+      "loss": 0.5066,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2116762430493864,
+      "learning_rate": 1.0606437737036849e-05,
+      "loss": 0.6093,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2845438509769538,
+      "learning_rate": 1.0603929671477669e-05,
+      "loss": 0.6411,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1755855055182531,
+      "learning_rate": 1.0601421567789289e-05,
+      "loss": 0.5427,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.14649404492435,
+      "learning_rate": 1.0598913426130067e-05,
+      "loss": 0.5598,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2899880525494083,
+      "learning_rate": 1.0596405246658348e-05,
+      "loss": 0.575,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.112014317310567,
+      "learning_rate": 1.0593897029532487e-05,
+      "loss": 0.5334,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0481667591163069,
+      "learning_rate": 1.0591388774910847e-05,
+      "loss": 0.544,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1072609328666656,
+      "learning_rate": 1.0588880482951778e-05,
+      "loss": 0.5097,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1279984886686873,
+      "learning_rate": 1.0586372153813649e-05,
+      "loss": 0.574,
+      "step": 6405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1469497905740156,
+      "learning_rate": 1.058386378765482e-05,
+      "loss": 0.5426,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.216356617252466,
+      "learning_rate": 1.0581355384633655e-05,
+      "loss": 0.5958,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1625914527442378,
+      "learning_rate": 1.0578846944908528e-05,
+      "loss": 0.4933,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1409076706208394,
+      "learning_rate": 1.0576338468637805e-05,
+      "loss": 0.591,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1989841293799026,
+      "learning_rate": 1.0573829955979864e-05,
+      "loss": 0.5128,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2182205713084933,
+      "learning_rate": 1.0571321407093076e-05,
+      "loss": 0.6106,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5133196414770653,
+      "learning_rate": 1.0568812822135819e-05,
+      "loss": 0.5609,
+      "step": 6412
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1435371742894647,
+      "learning_rate": 1.0566304201266473e-05,
+      "loss": 0.5382,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.212682488069999,
+      "learning_rate": 1.0563795544643422e-05,
+      "loss": 0.5654,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1751436044311612,
+      "learning_rate": 1.0561286852425052e-05,
+      "loss": 0.556,
+      "step": 6415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.297821159379831,
+      "learning_rate": 1.0558778124769747e-05,
+      "loss": 0.5435,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2000484516375873,
+      "learning_rate": 1.0556269361835891e-05,
+      "loss": 0.554,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1933590253713942,
+      "learning_rate": 1.0553760563781883e-05,
+      "loss": 0.5305,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1562032947901797,
+      "learning_rate": 1.0551251730766114e-05,
+      "loss": 0.5307,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3157086434131653,
+      "learning_rate": 1.054874286294698e-05,
+      "loss": 0.5819,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0745875609779298,
+      "learning_rate": 1.0546233960482876e-05,
+      "loss": 0.523,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.101834419182634,
+      "learning_rate": 1.0543725023532205e-05,
+      "loss": 0.5474,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2312347933996854,
+      "learning_rate": 1.0541216052253366e-05,
+      "loss": 0.5225,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.143481974647742,
+      "learning_rate": 1.0538707046804768e-05,
+      "loss": 0.5907,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3307627345838706,
+      "learning_rate": 1.0536198007344816e-05,
+      "loss": 0.5076,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1568975439610916,
+      "learning_rate": 1.0533688934031916e-05,
+      "loss": 0.5882,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1576403429377826,
+      "learning_rate": 1.0531179827024478e-05,
+      "loss": 0.5685,
+      "step": 6427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.322009694336566,
+      "learning_rate": 1.0528670686480918e-05,
+      "loss": 0.5823,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1246501060601453,
+      "learning_rate": 1.052616151255965e-05,
+      "loss": 0.5357,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1874247577320272,
+      "learning_rate": 1.0523652305419092e-05,
+      "loss": 0.5853,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2036285337522425,
+      "learning_rate": 1.0521143065217664e-05,
+      "loss": 0.6928,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0968589738071735,
+      "learning_rate": 1.0518633792113782e-05,
+      "loss": 0.5713,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1880031824647297,
+      "learning_rate": 1.051612448626587e-05,
+      "loss": 0.5931,
+      "step": 6433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1100371090811751,
+      "learning_rate": 1.0513615147832364e-05,
+      "loss": 0.525,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1496380609581494,
+      "learning_rate": 1.051110577697168e-05,
+      "loss": 0.607,
+      "step": 6435
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0676011876875386,
+      "learning_rate": 1.050859637384225e-05,
+      "loss": 0.5838,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0737795955145297,
+      "learning_rate": 1.050608693860251e-05,
+      "loss": 0.496,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1503433191906365,
+      "learning_rate": 1.0503577471410889e-05,
+      "loss": 0.5445,
+      "step": 6438
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1423082342667716,
+      "learning_rate": 1.0501067972425824e-05,
+      "loss": 0.5432,
+      "step": 6439
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1628046087885573,
+      "learning_rate": 1.0498558441805753e-05,
+      "loss": 0.5342,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.110299920296153,
+      "learning_rate": 1.0496048879709116e-05,
+      "loss": 0.5762,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.175374261622373,
+      "learning_rate": 1.0493539286294352e-05,
+      "loss": 0.5589,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1793212827448147,
+      "learning_rate": 1.0491029661719907e-05,
+      "loss": 0.5008,
+      "step": 6443
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.108561267016032,
+      "learning_rate": 1.0488520006144227e-05,
+      "loss": 0.5186,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2138178378789635,
+      "learning_rate": 1.0486010319725759e-05,
+      "loss": 0.6598,
+      "step": 6445
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1794840150876267,
+      "learning_rate": 1.0483500602622951e-05,
+      "loss": 0.5719,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1307855822480315,
+      "learning_rate": 1.0480990854994257e-05,
+      "loss": 0.542,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1702967888423044,
+      "learning_rate": 1.0478481076998127e-05,
+      "loss": 0.5754,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2183695957529324,
+      "learning_rate": 1.0475971268793019e-05,
+      "loss": 0.5705,
+      "step": 6449
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1231250503531536,
+      "learning_rate": 1.0473461430537388e-05,
+      "loss": 0.5615,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1772062974466484,
+      "learning_rate": 1.0470951562389695e-05,
+      "loss": 0.5459,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1594341869618447,
+      "learning_rate": 1.04684416645084e-05,
+      "loss": 0.5658,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1167754900747189,
+      "learning_rate": 1.0465931737051964e-05,
+      "loss": 0.5336,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1716011744699424,
+      "learning_rate": 1.0463421780178857e-05,
+      "loss": 0.5337,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2492031895644735,
+      "learning_rate": 1.0460911794047542e-05,
+      "loss": 0.5967,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.158638832002312,
+      "learning_rate": 1.0458401778816482e-05,
+      "loss": 0.5402,
+      "step": 6456
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.105887990653393,
+      "learning_rate": 1.0455891734644158e-05,
+      "loss": 0.5824,
+      "step": 6457
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2536555719994473,
+      "learning_rate": 1.0453381661689035e-05,
+      "loss": 0.5849,
+      "step": 6458
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2694543433736711,
+      "learning_rate": 1.045087156010959e-05,
+      "loss": 0.6392,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0417108272092714,
+      "learning_rate": 1.0448361430064296e-05,
+      "loss": 0.5086,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2602374472239826,
+      "learning_rate": 1.044585127171163e-05,
+      "loss": 0.5603,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1535407105278217,
+      "learning_rate": 1.0443341085210077e-05,
+      "loss": 0.535,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1858109458173658,
+      "learning_rate": 1.0440830870718108e-05,
+      "loss": 0.5779,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2357698599338334,
+      "learning_rate": 1.0438320628394219e-05,
+      "loss": 0.4938,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1479697280408172,
+      "learning_rate": 1.0435810358396882e-05,
+      "loss": 0.5555,
+      "step": 6465
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1795676972648603,
+      "learning_rate": 1.043330006088459e-05,
+      "loss": 0.5988,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2177928687459614,
+      "learning_rate": 1.0430789736015829e-05,
+      "loss": 0.6048,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1370087872973356,
+      "learning_rate": 1.0428279383949089e-05,
+      "loss": 0.5043,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.164688454093573,
+      "learning_rate": 1.0425769004842865e-05,
+      "loss": 0.5256,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2612197408836483,
+      "learning_rate": 1.0423258598855645e-05,
+      "loss": 0.5819,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1981013058989534,
+      "learning_rate": 1.0420748166145926e-05,
+      "loss": 0.6174,
+      "step": 6471
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3651490191625855,
+      "learning_rate": 1.0418237706872206e-05,
+      "loss": 0.596,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0885770265716346,
+      "learning_rate": 1.0415727221192977e-05,
+      "loss": 0.5212,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0616219202030865,
+      "learning_rate": 1.041321670926675e-05,
+      "loss": 0.4994,
+      "step": 6474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1637901141078357,
+      "learning_rate": 1.0410706171252017e-05,
+      "loss": 0.5353,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2454489352011935,
+      "learning_rate": 1.0408195607307283e-05,
+      "loss": 0.559,
+      "step": 6476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1409057376215426,
+      "learning_rate": 1.0405685017591057e-05,
+      "loss": 0.5341,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.17630221462805,
+      "learning_rate": 1.040317440226184e-05,
+      "loss": 0.5392,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2302056409702453,
+      "learning_rate": 1.0400663761478145e-05,
+      "loss": 0.5698,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1188492587080476,
+      "learning_rate": 1.039815309539848e-05,
+      "loss": 0.5897,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1265768021139977,
+      "learning_rate": 1.0395642404181355e-05,
+      "loss": 0.5319,
+      "step": 6481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2868430711621264,
+      "learning_rate": 1.0393131687985283e-05,
+      "loss": 0.5368,
+      "step": 6482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2184903773075373,
+      "learning_rate": 1.039062094696878e-05,
+      "loss": 0.5443,
+      "step": 6483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1159259101823455,
+      "learning_rate": 1.038811018129036e-05,
+      "loss": 0.4983,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1391438312562574,
+      "learning_rate": 1.0385599391108546e-05,
+      "loss": 0.5673,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1781463661583662,
+      "learning_rate": 1.0383088576581847e-05,
+      "loss": 0.5897,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1682227633546378,
+      "learning_rate": 1.0380577737868795e-05,
+      "loss": 0.5534,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0834568699067162,
+      "learning_rate": 1.0378066875127904e-05,
+      "loss": 0.5223,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.211989997495161,
+      "learning_rate": 1.03755559885177e-05,
+      "loss": 0.5408,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2030624274293904,
+      "learning_rate": 1.0373045078196713e-05,
+      "loss": 0.5529,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1346852574474686,
+      "learning_rate": 1.037053414432346e-05,
+      "loss": 0.5373,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9813056504814902,
+      "learning_rate": 1.0368023187056477e-05,
+      "loss": 0.4685,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.224564303552805,
+      "learning_rate": 1.0365512206554294e-05,
+      "loss": 0.5302,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2173315254374182,
+      "learning_rate": 1.0363001202975439e-05,
+      "loss": 0.5652,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1619547765142086,
+      "learning_rate": 1.0360490176478443e-05,
+      "loss": 0.5281,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3244909046216498,
+      "learning_rate": 1.0357979127221842e-05,
+      "loss": 0.618,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1800158214316623,
+      "learning_rate": 1.0355468055364171e-05,
+      "loss": 0.5515,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1716459940628152,
+      "learning_rate": 1.0352956961063972e-05,
+      "loss": 0.5988,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2110682755272661,
+      "learning_rate": 1.0350445844479775e-05,
+      "loss": 0.5478,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2359333100953198,
+      "learning_rate": 1.0347934705770126e-05,
+      "loss": 0.6416,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0897027474800642,
+      "learning_rate": 1.0345423545093563e-05,
+      "loss": 0.5335,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.228188135984859,
+      "learning_rate": 1.0342912362608628e-05,
+      "loss": 0.5646,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.122441667770039,
+      "learning_rate": 1.0340401158473869e-05,
+      "loss": 0.5241,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.106340314120338,
+      "learning_rate": 1.0337889932847828e-05,
+      "loss": 0.5667,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1356259366054169,
+      "learning_rate": 1.0335378685889053e-05,
+      "loss": 0.5444,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2017887671328018,
+      "learning_rate": 1.033286741775609e-05,
+      "loss": 0.537,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2231736231102452,
+      "learning_rate": 1.0330356128607489e-05,
+      "loss": 0.5794,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1158043361957681,
+      "learning_rate": 1.0327844818601802e-05,
+      "loss": 0.5099,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1514681835706855,
+      "learning_rate": 1.0325333487897579e-05,
+      "loss": 0.5287,
+      "step": 6509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2631800076050756,
+      "learning_rate": 1.0322822136653376e-05,
+      "loss": 0.5816,
+      "step": 6510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2516436737444387,
+      "learning_rate": 1.0320310765027746e-05,
+      "loss": 0.5551,
+      "step": 6511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0623838697808952,
+      "learning_rate": 1.0317799373179242e-05,
+      "loss": 0.5489,
+      "step": 6512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1943054444856593,
+      "learning_rate": 1.0315287961266427e-05,
+      "loss": 0.6022,
+      "step": 6513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1144732067332557,
+      "learning_rate": 1.0312776529447857e-05,
+      "loss": 0.5251,
+      "step": 6514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0348766021144762,
+      "learning_rate": 1.031026507788209e-05,
+      "loss": 0.5296,
+      "step": 6515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0781295333988135,
+      "learning_rate": 1.0307753606727685e-05,
+      "loss": 0.5198,
+      "step": 6516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2003042669791053,
+      "learning_rate": 1.0305242116143209e-05,
+      "loss": 0.537,
+      "step": 6517
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1016393864630574,
+      "learning_rate": 1.0302730606287226e-05,
+      "loss": 0.5368,
+      "step": 6518
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.153430594292088,
+      "learning_rate": 1.0300219077318294e-05,
+      "loss": 0.5304,
+      "step": 6519
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2458989580066322,
+      "learning_rate": 1.0297707529394984e-05,
+      "loss": 0.5492,
+      "step": 6520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1302176913153803,
+      "learning_rate": 1.0295195962675864e-05,
+      "loss": 0.5189,
+      "step": 6521
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1454132639442505,
+      "learning_rate": 1.0292684377319495e-05,
+      "loss": 0.5302,
+      "step": 6522
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1407558940874796,
+      "learning_rate": 1.0290172773484455e-05,
+      "loss": 0.5018,
+      "step": 6523
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0986633567802637,
+      "learning_rate": 1.0287661151329312e-05,
+      "loss": 0.5305,
+      "step": 6524
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.069648369561804,
+      "learning_rate": 1.0285149511012632e-05,
+      "loss": 0.5605,
+      "step": 6525
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2477564705169022,
+      "learning_rate": 1.0282637852692996e-05,
+      "loss": 0.5725,
+      "step": 6526
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2772713736931434,
+      "learning_rate": 1.028012617652897e-05,
+      "loss": 0.5525,
+      "step": 6527
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1024198509891407,
+      "learning_rate": 1.0277614482679136e-05,
+      "loss": 0.5512,
+      "step": 6528
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1801470436192658,
+      "learning_rate": 1.027510277130207e-05,
+      "loss": 0.5535,
+      "step": 6529
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1889222813063238,
+      "learning_rate": 1.027259104255634e-05,
+      "loss": 0.5709,
+      "step": 6530
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1782939334435116,
+      "learning_rate": 1.027007929660054e-05,
+      "loss": 0.5268,
+      "step": 6531
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1985508055758203,
+      "learning_rate": 1.0267567533593233e-05,
+      "loss": 0.5682,
+      "step": 6532
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0923754092476652,
+      "learning_rate": 1.026505575369301e-05,
+      "loss": 0.4851,
+      "step": 6533
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9899474862649464,
+      "learning_rate": 1.0262543957058451e-05,
+      "loss": 0.5371,
+      "step": 6534
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1415644917861016,
+      "learning_rate": 1.0260032143848134e-05,
+      "loss": 0.5193,
+      "step": 6535
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1370661358751857,
+      "learning_rate": 1.0257520314220647e-05,
+      "loss": 0.54,
+      "step": 6536
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0558569182361355,
+      "learning_rate": 1.0255008468334575e-05,
+      "loss": 0.5174,
+      "step": 6537
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1331114801945867,
+      "learning_rate": 1.0252496606348503e-05,
+      "loss": 0.5731,
+      "step": 6538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.216038009579673,
+      "learning_rate": 1.024998472842102e-05,
+      "loss": 0.5394,
+      "step": 6539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0205418047194605,
+      "learning_rate": 1.0247472834710708e-05,
+      "loss": 0.5132,
+      "step": 6540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1261061422619703,
+      "learning_rate": 1.024496092537616e-05,
+      "loss": 0.518,
+      "step": 6541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2768046790209342,
+      "learning_rate": 1.0242449000575963e-05,
+      "loss": 0.6223,
+      "step": 6542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.143730638684103,
+      "learning_rate": 1.0239937060468713e-05,
+      "loss": 0.5265,
+      "step": 6543
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.105533126196717,
+      "learning_rate": 1.0237425105212996e-05,
+      "loss": 0.5309,
+      "step": 6544
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1376822147515908,
+      "learning_rate": 1.0234913134967409e-05,
+      "loss": 0.5072,
+      "step": 6545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3780283956771728,
+      "learning_rate": 1.0232401149890544e-05,
+      "loss": 0.6105,
+      "step": 6546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1719564282419253,
+      "learning_rate": 1.0229889150140991e-05,
+      "loss": 0.5812,
+      "step": 6547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1801045167024367,
+      "learning_rate": 1.0227377135877354e-05,
+      "loss": 0.6039,
+      "step": 6548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3057707441253552,
+      "learning_rate": 1.0224865107258225e-05,
+      "loss": 0.5544,
+      "step": 6549
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.206555615385461,
+      "learning_rate": 1.0222353064442201e-05,
+      "loss": 0.5313,
+      "step": 6550
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1341920016726996,
+      "learning_rate": 1.0219841007587881e-05,
+      "loss": 0.4894,
+      "step": 6551
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1782227574272128,
+      "learning_rate": 1.0217328936853864e-05,
+      "loss": 0.5468,
+      "step": 6552
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0071137245799828,
+      "learning_rate": 1.021481685239875e-05,
+      "loss": 0.5132,
+      "step": 6553
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2673552189287915,
+      "learning_rate": 1.0212304754381139e-05,
+      "loss": 0.5952,
+      "step": 6554
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3368441087632277,
+      "learning_rate": 1.0209792642959634e-05,
+      "loss": 0.5348,
+      "step": 6555
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0995009048652482,
+      "learning_rate": 1.0207280518292837e-05,
+      "loss": 0.5244,
+      "step": 6556
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1372245371197143,
+      "learning_rate": 1.0204768380539352e-05,
+      "loss": 0.5116,
+      "step": 6557
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.155992015203863,
+      "learning_rate": 1.0202256229857781e-05,
+      "loss": 0.4794,
+      "step": 6558
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.254772184371548,
+      "learning_rate": 1.0199744066406735e-05,
+      "loss": 0.5829,
+      "step": 6559
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.09942245927465,
+      "learning_rate": 1.0197231890344809e-05,
+      "loss": 0.544,
+      "step": 6560
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2019369526575665,
+      "learning_rate": 1.0194719701830622e-05,
+      "loss": 0.5544,
+      "step": 6561
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.126106724490322,
+      "learning_rate": 1.0192207501022771e-05,
+      "loss": 0.5929,
+      "step": 6562
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2013904899826602,
+      "learning_rate": 1.0189695288079873e-05,
+      "loss": 0.5481,
+      "step": 6563
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2206449693178236,
+      "learning_rate": 1.0187183063160531e-05,
+      "loss": 0.6079,
+      "step": 6564
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.114337941664481,
+      "learning_rate": 1.0184670826423355e-05,
+      "loss": 0.5575,
+      "step": 6565
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2035699987463646,
+      "learning_rate": 1.018215857802696e-05,
+      "loss": 0.5743,
+      "step": 6566
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1147611185098123,
+      "learning_rate": 1.017964631812995e-05,
+      "loss": 0.552,
+      "step": 6567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2153862726131532,
+      "learning_rate": 1.0177134046890944e-05,
+      "loss": 0.5884,
+      "step": 6568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1977840626617977,
+      "learning_rate": 1.0174621764468553e-05,
+      "loss": 0.5469,
+      "step": 6569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0818145204384482,
+      "learning_rate": 1.0172109471021385e-05,
+      "loss": 0.5288,
+      "step": 6570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2089178655568527,
+      "learning_rate": 1.0169597166708061e-05,
+      "loss": 0.5069,
+      "step": 6571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1449184348693413,
+      "learning_rate": 1.0167084851687193e-05,
+      "loss": 0.5405,
+      "step": 6572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3064836484759563,
+      "learning_rate": 1.0164572526117396e-05,
+      "loss": 0.5443,
+      "step": 6573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2010027014219982,
+      "learning_rate": 1.0162060190157285e-05,
+      "loss": 0.5643,
+      "step": 6574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2232356054697529,
+      "learning_rate": 1.015954784396548e-05,
+      "loss": 0.5998,
+      "step": 6575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2996876102859292,
+      "learning_rate": 1.0157035487700592e-05,
+      "loss": 0.6302,
+      "step": 6576
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1514674588739016,
+      "learning_rate": 1.0154523121521249e-05,
+      "loss": 0.5674,
+      "step": 6577
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0995095785330056,
+      "learning_rate": 1.0152010745586064e-05,
+      "loss": 0.5496,
+      "step": 6578
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1704674978457041,
+      "learning_rate": 1.0149498360053656e-05,
+      "loss": 0.5491,
+      "step": 6579
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2343202892685265,
+      "learning_rate": 1.014698596508264e-05,
+      "loss": 0.5643,
+      "step": 6580
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1821688393580443,
+      "learning_rate": 1.0144473560831645e-05,
+      "loss": 0.5698,
+      "step": 6581
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1871477407699609,
+      "learning_rate": 1.0141961147459289e-05,
+      "loss": 0.5708,
+      "step": 6582
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1251248184473497,
+      "learning_rate": 1.0139448725124194e-05,
+      "loss": 0.5394,
+      "step": 6583
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2159984044547656,
+      "learning_rate": 1.0136936293984983e-05,
+      "loss": 0.5143,
+      "step": 6584
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1610665394676354,
+      "learning_rate": 1.0134423854200274e-05,
+      "loss": 0.5475,
+      "step": 6585
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0504687103266939,
+      "learning_rate": 1.0131911405928694e-05,
+      "loss": 0.4593,
+      "step": 6586
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0737200880047597,
+      "learning_rate": 1.0129398949328868e-05,
+      "loss": 0.5123,
+      "step": 6587
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.230050927325156,
+      "learning_rate": 1.0126886484559417e-05,
+      "loss": 0.5036,
+      "step": 6588
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.101732011021924,
+      "learning_rate": 1.012437401177897e-05,
+      "loss": 0.5352,
+      "step": 6589
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1523324222331206,
+      "learning_rate": 1.0121861531146147e-05,
+      "loss": 0.5841,
+      "step": 6590
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2120752219755637,
+      "learning_rate": 1.0119349042819578e-05,
+      "loss": 0.5782,
+      "step": 6591
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2562836542888172,
+      "learning_rate": 1.0116836546957891e-05,
+      "loss": 0.5675,
+      "step": 6592
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.190674203571873,
+      "learning_rate": 1.0114324043719705e-05,
+      "loss": 0.5856,
+      "step": 6593
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1906225910312582,
+      "learning_rate": 1.0111811533263656e-05,
+      "loss": 0.5644,
+      "step": 6594
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2795944217278874,
+      "learning_rate": 1.0109299015748364e-05,
+      "loss": 0.5661,
+      "step": 6595
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3449736944980009,
+      "learning_rate": 1.0106786491332464e-05,
+      "loss": 0.5914,
+      "step": 6596
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2152631225913308,
+      "learning_rate": 1.0104273960174584e-05,
+      "loss": 0.5965,
+      "step": 6597
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1265948964194838,
+      "learning_rate": 1.0101761422433348e-05,
+      "loss": 0.5187,
+      "step": 6598
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2651839253000516,
+      "learning_rate": 1.0099248878267387e-05,
+      "loss": 0.5677,
+      "step": 6599
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.195605971269191,
+      "learning_rate": 1.0096736327835335e-05,
+      "loss": 0.5369,
+      "step": 6600
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1175418105187145,
+      "learning_rate": 1.0094223771295817e-05,
+      "loss": 0.5727,
+      "step": 6601
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2082615973565014,
+      "learning_rate": 1.0091711208807471e-05,
+      "loss": 0.5648,
+      "step": 6602
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1560729638736833,
+      "learning_rate": 1.008919864052892e-05,
+      "loss": 0.5104,
+      "step": 6603
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1925774992545393,
+      "learning_rate": 1.0086686066618795e-05,
+      "loss": 0.5291,
+      "step": 6604
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1949278268606622,
+      "learning_rate": 1.0084173487235737e-05,
+      "loss": 0.6123,
+      "step": 6605
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.095876588197386,
+      "learning_rate": 1.008166090253837e-05,
+      "loss": 0.5442,
+      "step": 6606
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2621101268410695,
+      "learning_rate": 1.0079148312685334e-05,
+      "loss": 0.5848,
+      "step": 6607
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.172862997007255,
+      "learning_rate": 1.007663571783525e-05,
+      "loss": 0.5713,
+      "step": 6608
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1865492076194548,
+      "learning_rate": 1.007412311814676e-05,
+      "loss": 0.5255,
+      "step": 6609
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.195853815113784,
+      "learning_rate": 1.0071610513778494e-05,
+      "loss": 0.5771,
+      "step": 6610
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0966509359230607,
+      "learning_rate": 1.0069097904889087e-05,
+      "loss": 0.5368,
+      "step": 6611
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1946093916049472,
+      "learning_rate": 1.0066585291637176e-05,
+      "loss": 0.6046,
+      "step": 6612
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1394684558735477,
+      "learning_rate": 1.006407267418139e-05,
+      "loss": 0.5404,
+      "step": 6613
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1732227967134585,
+      "learning_rate": 1.0061560052680363e-05,
+      "loss": 0.5562,
+      "step": 6614
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1894899559895769,
+      "learning_rate": 1.0059047427292736e-05,
+      "loss": 0.5676,
+      "step": 6615
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2033830836535635,
+      "learning_rate": 1.0056534798177138e-05,
+      "loss": 0.5798,
+      "step": 6616
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2348429118958113,
+      "learning_rate": 1.005402216549221e-05,
+      "loss": 0.5515,
+      "step": 6617
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.103960173869838,
+      "learning_rate": 1.005150952939658e-05,
+      "loss": 0.5235,
+      "step": 6618
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1810135619881144,
+      "learning_rate": 1.0048996890048886e-05,
+      "loss": 0.527,
+      "step": 6619
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0970790336361398,
+      "learning_rate": 1.004648424760777e-05,
+      "loss": 0.4899,
+      "step": 6620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1840508211092986,
+      "learning_rate": 1.0043971602231862e-05,
+      "loss": 0.5785,
+      "step": 6621
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1611851198367176,
+      "learning_rate": 1.0041458954079801e-05,
+      "loss": 0.5741,
+      "step": 6622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.105816358499722,
+      "learning_rate": 1.003894630331022e-05,
+      "loss": 0.5021,
+      "step": 6623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1876195546250112,
+      "learning_rate": 1.0036433650081759e-05,
+      "loss": 0.518,
+      "step": 6624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.235508699939931,
+      "learning_rate": 1.0033920994553054e-05,
+      "loss": 0.5556,
+      "step": 6625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1087737267330544,
+      "learning_rate": 1.003140833688274e-05,
+      "loss": 0.4528,
+      "step": 6626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1096194897809544,
+      "learning_rate": 1.0028895677229458e-05,
+      "loss": 0.5449,
+      "step": 6627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.111613634170427,
+      "learning_rate": 1.0026383015751844e-05,
+      "loss": 0.5611,
+      "step": 6628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.209598071203132,
+      "learning_rate": 1.0023870352608529e-05,
+      "loss": 0.5912,
+      "step": 6629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1306951268485905,
+      "learning_rate": 1.002135768795816e-05,
+      "loss": 0.4876,
+      "step": 6630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1687027273970612,
+      "learning_rate": 1.0018845021959368e-05,
+      "loss": 0.5534,
+      "step": 6631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1685012573085252,
+      "learning_rate": 1.0016332354770792e-05,
+      "loss": 0.5429,
+      "step": 6632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1147397843598823,
+      "learning_rate": 1.0013819686551074e-05,
+      "loss": 0.5024,
+      "step": 6633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1629865141043234,
+      "learning_rate": 1.0011307017458843e-05,
+      "loss": 0.5447,
+      "step": 6634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1114728193164345,
+      "learning_rate": 1.0008794347652744e-05,
+      "loss": 0.5397,
+      "step": 6635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0910426282972538,
+      "learning_rate": 1.0006281677291411e-05,
+      "loss": 0.5077,
+      "step": 6636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.199703106312897,
+      "learning_rate": 1.0003769006533486e-05,
+      "loss": 0.5656,
+      "step": 6637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1136969074819794,
+      "learning_rate": 1.0001256335537604e-05,
+      "loss": 0.5386,
+      "step": 6638
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.123661887901538,
+      "learning_rate": 9.998743664462401e-06,
+      "loss": 0.5245,
+      "step": 6639
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1360052956941233,
+      "learning_rate": 9.996230993466517e-06,
+      "loss": 0.5185,
+      "step": 6640
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1896329095578912,
+      "learning_rate": 9.993718322708592e-06,
+      "loss": 0.5385,
+      "step": 6641
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1937608808251252,
+      "learning_rate": 9.99120565234726e-06,
+      "loss": 0.5775,
+      "step": 6642
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1417846532866414,
+      "learning_rate": 9.988692982541159e-06,
+      "loss": 0.5625,
+      "step": 6643
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2271778482631197,
+      "learning_rate": 9.986180313448933e-06,
+      "loss": 0.558,
+      "step": 6644
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1966580067235342,
+      "learning_rate": 9.98366764522921e-06,
+      "loss": 0.5977,
+      "step": 6645
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1479942867854638,
+      "learning_rate": 9.981154978040636e-06,
+      "loss": 0.5413,
+      "step": 6646
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1383323621303572,
+      "learning_rate": 9.978642312041843e-06,
+      "loss": 0.6055,
+      "step": 6647
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.176632899069099,
+      "learning_rate": 9.976129647391471e-06,
+      "loss": 0.5126,
+      "step": 6648
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1551517219305074,
+      "learning_rate": 9.973616984248162e-06,
+      "loss": 0.5925,
+      "step": 6649
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2088785202200736,
+      "learning_rate": 9.971104322770545e-06,
+      "loss": 0.6194,
+      "step": 6650
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1984971947953453,
+      "learning_rate": 9.968591663117263e-06,
+      "loss": 0.485,
+      "step": 6651
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1998027659293227,
+      "learning_rate": 9.966079005446949e-06,
+      "loss": 0.5658,
+      "step": 6652
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2112554317388005,
+      "learning_rate": 9.963566349918243e-06,
+      "loss": 0.5284,
+      "step": 6653
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0795774213265117,
+      "learning_rate": 9.961053696689785e-06,
+      "loss": 0.464,
+      "step": 6654
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3131545569076983,
+      "learning_rate": 9.958541045920204e-06,
+      "loss": 0.5983,
+      "step": 6655
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0439433141534018,
+      "learning_rate": 9.956028397768143e-06,
+      "loss": 0.5037,
+      "step": 6656
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0522319323655531,
+      "learning_rate": 9.953515752392233e-06,
+      "loss": 0.5325,
+      "step": 6657
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0810909110102431,
+      "learning_rate": 9.951003109951114e-06,
+      "loss": 0.5279,
+      "step": 6658
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.109893784982816,
+      "learning_rate": 9.948490470603425e-06,
+      "loss": 0.5258,
+      "step": 6659
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1533490467287377,
+      "learning_rate": 9.945977834507796e-06,
+      "loss": 0.5382,
+      "step": 6660
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0864777353272719,
+      "learning_rate": 9.943465201822865e-06,
+      "loss": 0.5156,
+      "step": 6661
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.116589899887985,
+      "learning_rate": 9.940952572707267e-06,
+      "loss": 0.5135,
+      "step": 6662
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1224579701700543,
+      "learning_rate": 9.938439947319639e-06,
+      "loss": 0.5442,
+      "step": 6663
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2587810601926352,
+      "learning_rate": 9.935927325818616e-06,
+      "loss": 0.5419,
+      "step": 6664
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0638526832876256,
+      "learning_rate": 9.933414708362829e-06,
+      "loss": 0.548,
+      "step": 6665
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3376120172812929,
+      "learning_rate": 9.930902095110916e-06,
+      "loss": 0.5877,
+      "step": 6666
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0784191891208117,
+      "learning_rate": 9.928389486221507e-06,
+      "loss": 0.49,
+      "step": 6667
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.232844695817907,
+      "learning_rate": 9.925876881853242e-06,
+      "loss": 0.5598,
+      "step": 6668
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2627350106898854,
+      "learning_rate": 9.92336428216475e-06,
+      "loss": 0.6332,
+      "step": 6669
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.175360670940941,
+      "learning_rate": 9.920851687314673e-06,
+      "loss": 0.5429,
+      "step": 6670
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2129608477966873,
+      "learning_rate": 9.918339097461631e-06,
+      "loss": 0.5597,
+      "step": 6671
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2944959397412539,
+      "learning_rate": 9.915826512764265e-06,
+      "loss": 0.5731,
+      "step": 6672
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1980466799740528,
+      "learning_rate": 9.913313933381203e-06,
+      "loss": 0.5604,
+      "step": 6673
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1586571972503563,
+      "learning_rate": 9.91080135947108e-06,
+      "loss": 0.5591,
+      "step": 6674
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2079291106038474,
+      "learning_rate": 9.908288791192532e-06,
+      "loss": 0.6064,
+      "step": 6675
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1589997046587932,
+      "learning_rate": 9.905776228704185e-06,
+      "loss": 0.5198,
+      "step": 6676
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0285197442273248,
+      "learning_rate": 9.903263672164668e-06,
+      "loss": 0.5242,
+      "step": 6677
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.135374657556637,
+      "learning_rate": 9.900751121732613e-06,
+      "loss": 0.5524,
+      "step": 6678
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1864209040542244,
+      "learning_rate": 9.898238577566654e-06,
+      "loss": 0.5433,
+      "step": 6679
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.193954943495299,
+      "learning_rate": 9.895726039825421e-06,
+      "loss": 0.5374,
+      "step": 6680
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1034898603014098,
+      "learning_rate": 9.89321350866754e-06,
+      "loss": 0.5038,
+      "step": 6681
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1904055035131913,
+      "learning_rate": 9.890700984251638e-06,
+      "loss": 0.5619,
+      "step": 6682
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0613637907671543,
+      "learning_rate": 9.888188466736347e-06,
+      "loss": 0.5205,
+      "step": 6683
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1402003921928328,
+      "learning_rate": 9.885675956280295e-06,
+      "loss": 0.5368,
+      "step": 6684
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2505179762526284,
+      "learning_rate": 9.883163453042115e-06,
+      "loss": 0.6086,
+      "step": 6685
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2176526827315106,
+      "learning_rate": 9.880650957180427e-06,
+      "loss": 0.5601,
+      "step": 6686
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1829621798547385,
+      "learning_rate": 9.878138468853856e-06,
+      "loss": 0.5446,
+      "step": 6687
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1524636933572054,
+      "learning_rate": 9.875625988221033e-06,
+      "loss": 0.5235,
+      "step": 6688
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.104295249265777,
+      "learning_rate": 9.873113515440583e-06,
+      "loss": 0.542,
+      "step": 6689
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1996496960823801,
+      "learning_rate": 9.870601050671137e-06,
+      "loss": 0.5884,
+      "step": 6690
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.158612852652886,
+      "learning_rate": 9.86808859407131e-06,
+      "loss": 0.5379,
+      "step": 6691
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2031040437532907,
+      "learning_rate": 9.865576145799729e-06,
+      "loss": 0.5425,
+      "step": 6692
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1316717784150023,
+      "learning_rate": 9.86306370601502e-06,
+      "loss": 0.517,
+      "step": 6693
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.325490470293655,
+      "learning_rate": 9.860551274875806e-06,
+      "loss": 0.5813,
+      "step": 6694
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2344710880013376,
+      "learning_rate": 9.858038852540713e-06,
+      "loss": 0.5207,
+      "step": 6695
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1529321752253516,
+      "learning_rate": 9.855526439168359e-06,
+      "loss": 0.5231,
+      "step": 6696
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1133799726649902,
+      "learning_rate": 9.853014034917361e-06,
+      "loss": 0.5354,
+      "step": 6697
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2422150422887546,
+      "learning_rate": 9.850501639946349e-06,
+      "loss": 0.5913,
+      "step": 6698
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2543223038397229,
+      "learning_rate": 9.847989254413938e-06,
+      "loss": 0.565,
+      "step": 6699
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.172135947102419,
+      "learning_rate": 9.845476878478754e-06,
+      "loss": 0.5242,
+      "step": 6700
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3321386389079346,
+      "learning_rate": 9.84296451229941e-06,
+      "loss": 0.608,
+      "step": 6701
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.127404557225205,
+      "learning_rate": 9.840452156034523e-06,
+      "loss": 0.5198,
+      "step": 6702
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.128997799022276,
+      "learning_rate": 9.837939809842717e-06,
+      "loss": 0.4907,
+      "step": 6703
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0543452519967167,
+      "learning_rate": 9.835427473882606e-06,
+      "loss": 0.5482,
+      "step": 6704
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1739258685808915,
+      "learning_rate": 9.83291514831281e-06,
+      "loss": 0.5725,
+      "step": 6705
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.18268084177563,
+      "learning_rate": 9.830402833291942e-06,
+      "loss": 0.5413,
+      "step": 6706
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1920730975895073,
+      "learning_rate": 9.827890528978617e-06,
+      "loss": 0.574,
+      "step": 6707
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1162062917697777,
+      "learning_rate": 9.82537823553145e-06,
+      "loss": 0.541,
+      "step": 6708
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2010820060543905,
+      "learning_rate": 9.822865953109055e-06,
+      "loss": 0.5556,
+      "step": 6709
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1876201568848452,
+      "learning_rate": 9.820353681870052e-06,
+      "loss": 0.5575,
+      "step": 6710
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1585664999310887,
+      "learning_rate": 9.817841421973046e-06,
+      "loss": 0.534,
+      "step": 6711
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1678580037051753,
+      "learning_rate": 9.815329173576648e-06,
+      "loss": 0.5664,
+      "step": 6712
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1033214296945408,
+      "learning_rate": 9.812816936839472e-06,
+      "loss": 0.528,
+      "step": 6713
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1152892090900546,
+      "learning_rate": 9.810304711920127e-06,
+      "loss": 0.5323,
+      "step": 6714
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1848273065334767,
+      "learning_rate": 9.80779249897723e-06,
+      "loss": 0.5652,
+      "step": 6715
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1234445415164733,
+      "learning_rate": 9.805280298169383e-06,
+      "loss": 0.562,
+      "step": 6716
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0949469964598548,
+      "learning_rate": 9.802768109655192e-06,
+      "loss": 0.5287,
+      "step": 6717
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.250375881900646,
+      "learning_rate": 9.800255933593269e-06,
+      "loss": 0.5525,
+      "step": 6718
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0301054758860864,
+      "learning_rate": 9.797743770142219e-06,
+      "loss": 0.5037,
+      "step": 6719
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1472910653846946,
+      "learning_rate": 9.795231619460652e-06,
+      "loss": 0.5301,
+      "step": 6720
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0766018598694587,
+      "learning_rate": 9.792719481707168e-06,
+      "loss": 0.4882,
+      "step": 6721
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1165987077122363,
+      "learning_rate": 9.79020735704037e-06,
+      "loss": 0.506,
+      "step": 6722
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0982313176959915,
+      "learning_rate": 9.787695245618864e-06,
+      "loss": 0.5402,
+      "step": 6723
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.104722974496296,
+      "learning_rate": 9.785183147601252e-06,
+      "loss": 0.534,
+      "step": 6724
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1942002850255082,
+      "learning_rate": 9.782671063146138e-06,
+      "loss": 0.5297,
+      "step": 6725
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2304632762000023,
+      "learning_rate": 9.780158992412124e-06,
+      "loss": 0.5548,
+      "step": 6726
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.038458110594458,
+      "learning_rate": 9.777646935557802e-06,
+      "loss": 0.5134,
+      "step": 6727
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2464058225166397,
+      "learning_rate": 9.775134892741778e-06,
+      "loss": 0.5645,
+      "step": 6728
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1757873332224278,
+      "learning_rate": 9.772622864122645e-06,
+      "loss": 0.5481,
+      "step": 6729
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0662611977994314,
+      "learning_rate": 9.770110849859009e-06,
+      "loss": 0.5524,
+      "step": 6730
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1548211832051936,
+      "learning_rate": 9.767598850109463e-06,
+      "loss": 0.5567,
+      "step": 6731
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1617516231439333,
+      "learning_rate": 9.765086865032596e-06,
+      "loss": 0.4977,
+      "step": 6732
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3047199359448107,
+      "learning_rate": 9.762574894787006e-06,
+      "loss": 0.5427,
+      "step": 6733
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.203732126802435,
+      "learning_rate": 9.760062939531289e-06,
+      "loss": 0.5459,
+      "step": 6734
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1284393084038036,
+      "learning_rate": 9.757550999424038e-06,
+      "loss": 0.5466,
+      "step": 6735
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1869552266193888,
+      "learning_rate": 9.755039074623846e-06,
+      "loss": 0.5742,
+      "step": 6736
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.632009865642161,
+      "learning_rate": 9.752527165289297e-06,
+      "loss": 0.6006,
+      "step": 6737
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2485300480958519,
+      "learning_rate": 9.750015271578982e-06,
+      "loss": 0.5613,
+      "step": 6738
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1568116035587794,
+      "learning_rate": 9.747503393651499e-06,
+      "loss": 0.5488,
+      "step": 6739
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0118085313234109,
+      "learning_rate": 9.744991531665425e-06,
+      "loss": 0.5363,
+      "step": 6740
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1239115960152977,
+      "learning_rate": 9.742479685779356e-06,
+      "loss": 0.5595,
+      "step": 6741
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.262332306479604,
+      "learning_rate": 9.739967856151868e-06,
+      "loss": 0.5743,
+      "step": 6742
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.251172660091367,
+      "learning_rate": 9.737456042941552e-06,
+      "loss": 0.5425,
+      "step": 6743
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.156354538599304,
+      "learning_rate": 9.734944246306994e-06,
+      "loss": 0.4942,
+      "step": 6744
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2061475463208424,
+      "learning_rate": 9.732432466406769e-06,
+      "loss": 0.5433,
+      "step": 6745
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1490041152442163,
+      "learning_rate": 9.729920703399468e-06,
+      "loss": 0.5294,
+      "step": 6746
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.170693730395212,
+      "learning_rate": 9.727408957443661e-06,
+      "loss": 0.5873,
+      "step": 6747
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1811728316032652,
+      "learning_rate": 9.724897228697933e-06,
+      "loss": 0.5533,
+      "step": 6748
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2222789633228246,
+      "learning_rate": 9.722385517320866e-06,
+      "loss": 0.5175,
+      "step": 6749
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0903614960997197,
+      "learning_rate": 9.71987382347103e-06,
+      "loss": 0.5097,
+      "step": 6750
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.103902455274103,
+      "learning_rate": 9.717362147307009e-06,
+      "loss": 0.5288,
+      "step": 6751
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1689437827958349,
+      "learning_rate": 9.714850488987371e-06,
+      "loss": 0.531,
+      "step": 6752
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1734801425627166,
+      "learning_rate": 9.712338848670691e-06,
+      "loss": 0.5627,
+      "step": 6753
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.108512381860031,
+      "learning_rate": 9.709827226515547e-06,
+      "loss": 0.4915,
+      "step": 6754
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0919641356395131,
+      "learning_rate": 9.707315622680505e-06,
+      "loss": 0.4829,
+      "step": 6755
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.157064923936836,
+      "learning_rate": 9.70480403732414e-06,
+      "loss": 0.5811,
+      "step": 6756
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.182154419224635,
+      "learning_rate": 9.702292470605017e-06,
+      "loss": 0.562,
+      "step": 6757
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1521575260014056,
+      "learning_rate": 9.699780922681707e-06,
+      "loss": 0.6059,
+      "step": 6758
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1294052446766776,
+      "learning_rate": 9.697269393712779e-06,
+      "loss": 0.5141,
+      "step": 6759
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1084767318162467,
+      "learning_rate": 9.694757883856791e-06,
+      "loss": 0.4807,
+      "step": 6760
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.081503012717782,
+      "learning_rate": 9.69224639327232e-06,
+      "loss": 0.5312,
+      "step": 6761
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1142215973621492,
+      "learning_rate": 9.689734922117915e-06,
+      "loss": 0.4868,
+      "step": 6762
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1839190750464534,
+      "learning_rate": 9.687223470552146e-06,
+      "loss": 0.545,
+      "step": 6763
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2167807588688702,
+      "learning_rate": 9.684712038733575e-06,
+      "loss": 0.547,
+      "step": 6764
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1254207036382493,
+      "learning_rate": 9.682200626820758e-06,
+      "loss": 0.5596,
+      "step": 6765
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1276042253994945,
+      "learning_rate": 9.679689234972259e-06,
+      "loss": 0.5294,
+      "step": 6766
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1455863802113557,
+      "learning_rate": 9.677177863346627e-06,
+      "loss": 0.5391,
+      "step": 6767
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1438606041965453,
+      "learning_rate": 9.674666512102423e-06,
+      "loss": 0.4781,
+      "step": 6768
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.190637609460096,
+      "learning_rate": 9.672155181398201e-06,
+      "loss": 0.5484,
+      "step": 6769
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0915448438700075,
+      "learning_rate": 9.669643871392513e-06,
+      "loss": 0.4701,
+      "step": 6770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.32897414942817,
+      "learning_rate": 9.667132582243916e-06,
+      "loss": 0.5755,
+      "step": 6771
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2611725277353434,
+      "learning_rate": 9.664621314110952e-06,
+      "loss": 0.5244,
+      "step": 6772
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.109609498505618,
+      "learning_rate": 9.662110067152173e-06,
+      "loss": 0.4966,
+      "step": 6773
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1127352797923649,
+      "learning_rate": 9.659598841526133e-06,
+      "loss": 0.5074,
+      "step": 6774
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0763404561840126,
+      "learning_rate": 9.657087637391372e-06,
+      "loss": 0.4972,
+      "step": 6775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1935854134266628,
+      "learning_rate": 9.654576454906437e-06,
+      "loss": 0.5213,
+      "step": 6776
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2369850667177,
+      "learning_rate": 9.652065294229877e-06,
+      "loss": 0.5326,
+      "step": 6777
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1852745987826399,
+      "learning_rate": 9.649554155520227e-06,
+      "loss": 0.5173,
+      "step": 6778
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0512529753689384,
+      "learning_rate": 9.647043038936033e-06,
+      "loss": 0.5297,
+      "step": 6779
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1237311625360409,
+      "learning_rate": 9.644531944635829e-06,
+      "loss": 0.5414,
+      "step": 6780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3321954172942962,
+      "learning_rate": 9.64202087277816e-06,
+      "loss": 0.604,
+      "step": 6781
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.098598361004965,
+      "learning_rate": 9.639509823521562e-06,
+      "loss": 0.5335,
+      "step": 6782
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2298731740059234,
+      "learning_rate": 9.636998797024566e-06,
+      "loss": 0.5467,
+      "step": 6783
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.092405582861112,
+      "learning_rate": 9.634487793445711e-06,
+      "loss": 0.552,
+      "step": 6784
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0590108352445429,
+      "learning_rate": 9.631976812943523e-06,
+      "loss": 0.468,
+      "step": 6785
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1622103607153123,
+      "learning_rate": 9.62946585567654e-06,
+      "loss": 0.5303,
+      "step": 6786
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2147668648581655,
+      "learning_rate": 9.626954921803294e-06,
+      "loss": 0.5373,
+      "step": 6787
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1345811913944404,
+      "learning_rate": 9.624444011482302e-06,
+      "loss": 0.5027,
+      "step": 6788
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3148610812590216,
+      "learning_rate": 9.6219331248721e-06,
+      "loss": 0.6111,
+      "step": 6789
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.120634775879702,
+      "learning_rate": 9.619422262131208e-06,
+      "loss": 0.4998,
+      "step": 6790
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.18938006096794,
+      "learning_rate": 9.616911423418153e-06,
+      "loss": 0.5491,
+      "step": 6791
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3028522764096522,
+      "learning_rate": 9.61440060889146e-06,
+      "loss": 0.6272,
+      "step": 6792
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0668678262890638,
+      "learning_rate": 9.611889818709643e-06,
+      "loss": 0.5332,
+      "step": 6793
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1850208202577182,
+      "learning_rate": 9.609379053031224e-06,
+      "loss": 0.5584,
+      "step": 6794
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1292958362829233,
+      "learning_rate": 9.60686831201472e-06,
+      "loss": 0.5175,
+      "step": 6795
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1613028155615366,
+      "learning_rate": 9.604357595818647e-06,
+      "loss": 0.5409,
+      "step": 6796
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1523969217737962,
+      "learning_rate": 9.601846904601526e-06,
+      "loss": 0.608,
+      "step": 6797
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1970531779198983,
+      "learning_rate": 9.599336238521859e-06,
+      "loss": 0.5111,
+      "step": 6798
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1778648898810051,
+      "learning_rate": 9.596825597738164e-06,
+      "loss": 0.5495,
+      "step": 6799
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1870283896916536,
+      "learning_rate": 9.594314982408947e-06,
+      "loss": 0.5553,
+      "step": 6800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1901931840700481,
+      "learning_rate": 9.591804392692719e-06,
+      "loss": 0.5627,
+      "step": 6801
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1053506093874272,
+      "learning_rate": 9.589293828747988e-06,
+      "loss": 0.5174,
+      "step": 6802
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2732401969977203,
+      "learning_rate": 9.586783290733254e-06,
+      "loss": 0.5677,
+      "step": 6803
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1656762528593017,
+      "learning_rate": 9.584272778807026e-06,
+      "loss": 0.5585,
+      "step": 6804
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1832376580222328,
+      "learning_rate": 9.581762293127798e-06,
+      "loss": 0.5162,
+      "step": 6805
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0895200952932398,
+      "learning_rate": 9.579251833854076e-06,
+      "loss": 0.5438,
+      "step": 6806
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.126061786162506,
+      "learning_rate": 9.57674140114436e-06,
+      "loss": 0.5528,
+      "step": 6807
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2020671206504352,
+      "learning_rate": 9.574230995157137e-06,
+      "loss": 0.5021,
+      "step": 6808
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2667014642168681,
+      "learning_rate": 9.571720616050913e-06,
+      "loss": 0.5896,
+      "step": 6809
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1196191107699502,
+      "learning_rate": 9.569210263984173e-06,
+      "loss": 0.517,
+      "step": 6810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0555079906744835,
+      "learning_rate": 9.566699939115412e-06,
+      "loss": 0.5122,
+      "step": 6811
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1921523962689797,
+      "learning_rate": 9.564189641603123e-06,
+      "loss": 0.5579,
+      "step": 6812
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2114677498807633,
+      "learning_rate": 9.561679371605786e-06,
+      "loss": 0.5154,
+      "step": 6813
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.132669584532855,
+      "learning_rate": 9.559169129281893e-06,
+      "loss": 0.5237,
+      "step": 6814
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1251716482822274,
+      "learning_rate": 9.556658914789926e-06,
+      "loss": 0.5107,
+      "step": 6815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1155361962971095,
+      "learning_rate": 9.554148728288371e-06,
+      "loss": 0.488,
+      "step": 6816
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2365542622316654,
+      "learning_rate": 9.551638569935708e-06,
+      "loss": 0.5538,
+      "step": 6817
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1413754124167022,
+      "learning_rate": 9.549128439890413e-06,
+      "loss": 0.5174,
+      "step": 6818
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.210704461100857,
+      "learning_rate": 9.546618338310968e-06,
+      "loss": 0.5515,
+      "step": 6819
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.158660643917127,
+      "learning_rate": 9.544108265355843e-06,
+      "loss": 0.5085,
+      "step": 6820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2350867911201326,
+      "learning_rate": 9.541598221183516e-06,
+      "loss": 0.5753,
+      "step": 6821
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.192111247612642,
+      "learning_rate": 9.539088205952463e-06,
+      "loss": 0.4856,
+      "step": 6822
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1520995835071648,
+      "learning_rate": 9.536578219821146e-06,
+      "loss": 0.4988,
+      "step": 6823
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.298774477202078,
+      "learning_rate": 9.53406826294804e-06,
+      "loss": 0.5703,
+      "step": 6824
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.22550797537454,
+      "learning_rate": 9.531558335491602e-06,
+      "loss": 0.531,
+      "step": 6825
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.168792587157156,
+      "learning_rate": 9.529048437610307e-06,
+      "loss": 0.5501,
+      "step": 6826
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2242872675434198,
+      "learning_rate": 9.526538569462617e-06,
+      "loss": 0.5519,
+      "step": 6827
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2389060288447225,
+      "learning_rate": 9.524028731206984e-06,
+      "loss": 0.5388,
+      "step": 6828
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1037690808384706,
+      "learning_rate": 9.521518923001877e-06,
+      "loss": 0.446,
+      "step": 6829
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1709760651296448,
+      "learning_rate": 9.519009145005747e-06,
+      "loss": 0.5549,
+      "step": 6830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1214836585648127,
+      "learning_rate": 9.516499397377049e-06,
+      "loss": 0.5331,
+      "step": 6831
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1658898673939344,
+      "learning_rate": 9.513989680274241e-06,
+      "loss": 0.5721,
+      "step": 6832
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1407743904838614,
+      "learning_rate": 9.511479993855776e-06,
+      "loss": 0.5538,
+      "step": 6833
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2103429011302314,
+      "learning_rate": 9.508970338280097e-06,
+      "loss": 0.5871,
+      "step": 6834
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1783403194783333,
+      "learning_rate": 9.50646071370565e-06,
+      "loss": 0.563,
+      "step": 6835
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.25163257797966,
+      "learning_rate": 9.503951120290886e-06,
+      "loss": 0.5629,
+      "step": 6836
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0621521324202032,
+      "learning_rate": 9.501441558194247e-06,
+      "loss": 0.5365,
+      "step": 6837
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3286018413147922,
+      "learning_rate": 9.498932027574179e-06,
+      "loss": 0.5943,
+      "step": 6838
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.140774860727318,
+      "learning_rate": 9.496422528589115e-06,
+      "loss": 0.5767,
+      "step": 6839
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2509644125855384,
+      "learning_rate": 9.493913061397493e-06,
+      "loss": 0.5278,
+      "step": 6840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0820306440551384,
+      "learning_rate": 9.49140362615775e-06,
+      "loss": 0.5561,
+      "step": 6841
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1487220359777455,
+      "learning_rate": 9.488894223028322e-06,
+      "loss": 0.5719,
+      "step": 6842
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2308671578778951,
+      "learning_rate": 9.48638485216764e-06,
+      "loss": 0.5494,
+      "step": 6843
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1264529382599893,
+      "learning_rate": 9.483875513734131e-06,
+      "loss": 0.5611,
+      "step": 6844
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1681477081067346,
+      "learning_rate": 9.481366207886223e-06,
+      "loss": 0.5515,
+      "step": 6845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1517450878882316,
+      "learning_rate": 9.47885693478234e-06,
+      "loss": 0.5623,
+      "step": 6846
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1045334164912133,
+      "learning_rate": 9.476347694580911e-06,
+      "loss": 0.5322,
+      "step": 6847
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.156335363587699,
+      "learning_rate": 9.473838487440354e-06,
+      "loss": 0.5257,
+      "step": 6848
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1668430376706829,
+      "learning_rate": 9.471329313519086e-06,
+      "loss": 0.5719,
+      "step": 6849
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1390214910045442,
+      "learning_rate": 9.468820172975525e-06,
+      "loss": 0.5465,
+      "step": 6850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3029624361872645,
+      "learning_rate": 9.466311065968088e-06,
+      "loss": 0.5726,
+      "step": 6851
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0708037870811096,
+      "learning_rate": 9.463801992655187e-06,
+      "loss": 0.5075,
+      "step": 6852
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1286208376062532,
+      "learning_rate": 9.461292953195234e-06,
+      "loss": 0.5303,
+      "step": 6853
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1378655695096467,
+      "learning_rate": 9.458783947746635e-06,
+      "loss": 0.4978,
+      "step": 6854
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2296301027350895,
+      "learning_rate": 9.456274976467798e-06,
+      "loss": 0.5611,
+      "step": 6855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2008565845572483,
+      "learning_rate": 9.453766039517126e-06,
+      "loss": 0.5897,
+      "step": 6856
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1358637789766757,
+      "learning_rate": 9.451257137053022e-06,
+      "loss": 0.5175,
+      "step": 6857
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0870034454417437,
+      "learning_rate": 9.448748269233889e-06,
+      "loss": 0.5717,
+      "step": 6858
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1680117698069887,
+      "learning_rate": 9.44623943621812e-06,
+      "loss": 0.562,
+      "step": 6859
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2131268797933095,
+      "learning_rate": 9.443730638164112e-06,
+      "loss": 0.5597,
+      "step": 6860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.161619195218998,
+      "learning_rate": 9.441221875230257e-06,
+      "loss": 0.5318,
+      "step": 6861
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.193824889545189,
+      "learning_rate": 9.438713147574951e-06,
+      "loss": 0.5891,
+      "step": 6862
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2061538717281686,
+      "learning_rate": 9.43620445535658e-06,
+      "loss": 0.5521,
+      "step": 6863
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1358672423326084,
+      "learning_rate": 9.43369579873353e-06,
+      "loss": 0.5489,
+      "step": 6864
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.112326014768923,
+      "learning_rate": 9.431187177864184e-06,
+      "loss": 0.5347,
+      "step": 6865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2957539596452956,
+      "learning_rate": 9.428678592906925e-06,
+      "loss": 0.5118,
+      "step": 6866
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2199467383155636,
+      "learning_rate": 9.426170044020139e-06,
+      "loss": 0.5734,
+      "step": 6867
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2289500234171722,
+      "learning_rate": 9.423661531362197e-06,
+      "loss": 0.5753,
+      "step": 6868
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.159313576605824,
+      "learning_rate": 9.421153055091477e-06,
+      "loss": 0.5575,
+      "step": 6869
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.133401967289168,
+      "learning_rate": 9.418644615366346e-06,
+      "loss": 0.5535,
+      "step": 6870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0925077195407658,
+      "learning_rate": 9.416136212345183e-06,
+      "loss": 0.5117,
+      "step": 6871
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1616052383553008,
+      "learning_rate": 9.413627846186354e-06,
+      "loss": 0.5611,
+      "step": 6872
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.192798388994563,
+      "learning_rate": 9.411119517048226e-06,
+      "loss": 0.5265,
+      "step": 6873
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.172370195666303,
+      "learning_rate": 9.40861122508916e-06,
+      "loss": 0.5269,
+      "step": 6874
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.13235438061179,
+      "learning_rate": 9.406102970467515e-06,
+      "loss": 0.5494,
+      "step": 6875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1302768082977968,
+      "learning_rate": 9.403594753341655e-06,
+      "loss": 0.5299,
+      "step": 6876
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1358724898214763,
+      "learning_rate": 9.401086573869937e-06,
+      "loss": 0.5687,
+      "step": 6877
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1984956033440617,
+      "learning_rate": 9.398578432210713e-06,
+      "loss": 0.5353,
+      "step": 6878
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.262870664288372,
+      "learning_rate": 9.396070328522336e-06,
+      "loss": 0.5797,
+      "step": 6879
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2578837688466031,
+      "learning_rate": 9.393562262963154e-06,
+      "loss": 0.5696,
+      "step": 6880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0445816769692662,
+      "learning_rate": 9.391054235691515e-06,
+      "loss": 0.4845,
+      "step": 6881
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1939792552724164,
+      "learning_rate": 9.388546246865764e-06,
+      "loss": 0.5524,
+      "step": 6882
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.185410267031021,
+      "learning_rate": 9.386038296644245e-06,
+      "loss": 0.5581,
+      "step": 6883
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1838205957358947,
+      "learning_rate": 9.383530385185295e-06,
+      "loss": 0.5594,
+      "step": 6884
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2021331662184975,
+      "learning_rate": 9.381022512647251e-06,
+      "loss": 0.5972,
+      "step": 6885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2612453552856213,
+      "learning_rate": 9.37851467918845e-06,
+      "loss": 0.5922,
+      "step": 6886
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.115468977588695,
+      "learning_rate": 9.376006884967224e-06,
+      "loss": 0.473,
+      "step": 6887
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.140491111944837,
+      "learning_rate": 9.373499130141901e-06,
+      "loss": 0.5621,
+      "step": 6888
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0999414948597301,
+      "learning_rate": 9.370991414870814e-06,
+      "loss": 0.5277,
+      "step": 6889
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1706286605557767,
+      "learning_rate": 9.368483739312281e-06,
+      "loss": 0.5493,
+      "step": 6890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1421705252538015,
+      "learning_rate": 9.365976103624628e-06,
+      "loss": 0.554,
+      "step": 6891
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.174512007001139,
+      "learning_rate": 9.363468507966175e-06,
+      "loss": 0.5224,
+      "step": 6892
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1151110151645698,
+      "learning_rate": 9.360960952495239e-06,
+      "loss": 0.5403,
+      "step": 6893
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2198132498656216,
+      "learning_rate": 9.35845343737014e-06,
+      "loss": 0.5749,
+      "step": 6894
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2073646751536584,
+      "learning_rate": 9.355945962749179e-06,
+      "loss": 0.5631,
+      "step": 6895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1083731627418414,
+      "learning_rate": 9.353438528790673e-06,
+      "loss": 0.4935,
+      "step": 6896
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1268531054138002,
+      "learning_rate": 9.350931135652932e-06,
+      "loss": 0.5463,
+      "step": 6897
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2076929741427251,
+      "learning_rate": 9.348423783494253e-06,
+      "loss": 0.5921,
+      "step": 6898
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1776943421400163,
+      "learning_rate": 9.345916472472947e-06,
+      "loss": 0.5425,
+      "step": 6899
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2562159481462871,
+      "learning_rate": 9.343409202747306e-06,
+      "loss": 0.591,
+      "step": 6900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0744898228649296,
+      "learning_rate": 9.340901974475627e-06,
+      "loss": 0.5436,
+      "step": 6901
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2216132837387694,
+      "learning_rate": 9.33839478781621e-06,
+      "loss": 0.5736,
+      "step": 6902
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1346967088582525,
+      "learning_rate": 9.335887642927342e-06,
+      "loss": 0.5508,
+      "step": 6903
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2722628271618521,
+      "learning_rate": 9.333380539967319e-06,
+      "loss": 0.5887,
+      "step": 6904
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1576385408546024,
+      "learning_rate": 9.330873479094415e-06,
+      "loss": 0.5157,
+      "step": 6905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0669097829466203,
+      "learning_rate": 9.328366460466921e-06,
+      "loss": 0.5398,
+      "step": 6906
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.140139228115547,
+      "learning_rate": 9.325859484243121e-06,
+      "loss": 0.552,
+      "step": 6907
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1109227948080684,
+      "learning_rate": 9.323352550581285e-06,
+      "loss": 0.4995,
+      "step": 6908
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.323131688983571,
+      "learning_rate": 9.320845659639702e-06,
+      "loss": 0.6106,
+      "step": 6909
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1660169540729968,
+      "learning_rate": 9.31833881157663e-06,
+      "loss": 0.5586,
+      "step": 6910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2727689802005524,
+      "learning_rate": 9.315832006550345e-06,
+      "loss": 0.5651,
+      "step": 6911
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2606625698015634,
+      "learning_rate": 9.313325244719117e-06,
+      "loss": 0.5756,
+      "step": 6912
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1544550805024634,
+      "learning_rate": 9.310818526241209e-06,
+      "loss": 0.515,
+      "step": 6913
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2779881896128449,
+      "learning_rate": 9.308311851274885e-06,
+      "loss": 0.5994,
+      "step": 6914
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2377965321386084,
+      "learning_rate": 9.3058052199784e-06,
+      "loss": 0.5549,
+      "step": 6915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1371878478617607,
+      "learning_rate": 9.303298632510014e-06,
+      "loss": 0.537,
+      "step": 6916
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.149749162518121,
+      "learning_rate": 9.30079208902798e-06,
+      "loss": 0.5475,
+      "step": 6917
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3233202471354304,
+      "learning_rate": 9.298285589690548e-06,
+      "loss": 0.5925,
+      "step": 6918
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.063030951733851,
+      "learning_rate": 9.29577913465597e-06,
+      "loss": 0.5224,
+      "step": 6919
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1598342808374553,
+      "learning_rate": 9.293272724082484e-06,
+      "loss": 0.5138,
+      "step": 6920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.150487044071225,
+      "learning_rate": 9.290766358128338e-06,
+      "loss": 0.4975,
+      "step": 6921
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1169438129695117,
+      "learning_rate": 9.288260036951774e-06,
+      "loss": 0.5267,
+      "step": 6922
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2041056216033064,
+      "learning_rate": 9.285753760711023e-06,
+      "loss": 0.5532,
+      "step": 6923
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1674193043563388,
+      "learning_rate": 9.283247529564326e-06,
+      "loss": 0.5646,
+      "step": 6924
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1768027395955605,
+      "learning_rate": 9.280741343669908e-06,
+      "loss": 0.545,
+      "step": 6925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.081877494381608,
+      "learning_rate": 9.278235203185999e-06,
+      "loss": 0.4693,
+      "step": 6926
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1036741969862582,
+      "learning_rate": 9.275729108270825e-06,
+      "loss": 0.5205,
+      "step": 6927
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1665832058254102,
+      "learning_rate": 9.27322305908261e-06,
+      "loss": 0.4809,
+      "step": 6928
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2345403246684257,
+      "learning_rate": 9.270717055779575e-06,
+      "loss": 0.5734,
+      "step": 6929
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1085644837079052,
+      "learning_rate": 9.268211098519932e-06,
+      "loss": 0.4978,
+      "step": 6930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0934605351649394,
+      "learning_rate": 9.265705187461898e-06,
+      "loss": 0.5376,
+      "step": 6931
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.117967133484211,
+      "learning_rate": 9.263199322763687e-06,
+      "loss": 0.5437,
+      "step": 6932
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1633346125223145,
+      "learning_rate": 9.2606935045835e-06,
+      "loss": 0.5395,
+      "step": 6933
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2807320268849065,
+      "learning_rate": 9.258187733079552e-06,
+      "loss": 0.5728,
+      "step": 6934
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1802597173877796,
+      "learning_rate": 9.255682008410034e-06,
+      "loss": 0.5462,
+      "step": 6935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0465943828657278,
+      "learning_rate": 9.253176330733153e-06,
+      "loss": 0.4721,
+      "step": 6936
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0977897698589962,
+      "learning_rate": 9.250670700207103e-06,
+      "loss": 0.4856,
+      "step": 6937
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1913492658151363,
+      "learning_rate": 9.248165116990078e-06,
+      "loss": 0.5302,
+      "step": 6938
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2471735470191296,
+      "learning_rate": 9.245659581240271e-06,
+      "loss": 0.5617,
+      "step": 6939
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0122175128723463,
+      "learning_rate": 9.243154093115862e-06,
+      "loss": 0.4811,
+      "step": 6940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1170565121561231,
+      "learning_rate": 9.240648652775042e-06,
+      "loss": 0.5079,
+      "step": 6941
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2224205691816439,
+      "learning_rate": 9.23814326037599e-06,
+      "loss": 0.5465,
+      "step": 6942
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1123657209817601,
+      "learning_rate": 9.235637916076884e-06,
+      "loss": 0.5586,
+      "step": 6943
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0958176823261065,
+      "learning_rate": 9.233132620035899e-06,
+      "loss": 0.5004,
+      "step": 6944
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1805004830074655,
+      "learning_rate": 9.230627372411213e-06,
+      "loss": 0.5226,
+      "step": 6945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1487981009556634,
+      "learning_rate": 9.228122173360986e-06,
+      "loss": 0.4874,
+      "step": 6946
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1204726462508574,
+      "learning_rate": 9.22561702304339e-06,
+      "loss": 0.5522,
+      "step": 6947
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.189242941039336,
+      "learning_rate": 9.223111921616585e-06,
+      "loss": 0.5627,
+      "step": 6948
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.184960108540278,
+      "learning_rate": 9.220606869238733e-06,
+      "loss": 0.5342,
+      "step": 6949
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2431689529742278,
+      "learning_rate": 9.218101866067995e-06,
+      "loss": 0.5962,
+      "step": 6950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1778757696852609,
+      "learning_rate": 9.215596912262515e-06,
+      "loss": 0.5381,
+      "step": 6951
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2069500645861622,
+      "learning_rate": 9.213092007980453e-06,
+      "loss": 0.5376,
+      "step": 6952
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.151024324910753,
+      "learning_rate": 9.21058715337995e-06,
+      "loss": 0.5071,
+      "step": 6953
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0645189177452357,
+      "learning_rate": 9.208082348619149e-06,
+      "loss": 0.5426,
+      "step": 6954
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1235549440925061,
+      "learning_rate": 9.205577593856203e-06,
+      "loss": 0.5517,
+      "step": 6955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1496807297893243,
+      "learning_rate": 9.203072889249237e-06,
+      "loss": 0.5198,
+      "step": 6956
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0767754110292076,
+      "learning_rate": 9.200568234956393e-06,
+      "loss": 0.5457,
+      "step": 6957
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1517957515849504,
+      "learning_rate": 9.198063631135799e-06,
+      "loss": 0.5566,
+      "step": 6958
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.173953184552595,
+      "learning_rate": 9.195559077945586e-06,
+      "loss": 0.564,
+      "step": 6959
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.221985215389513,
+      "learning_rate": 9.193054575543883e-06,
+      "loss": 0.5874,
+      "step": 6960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2115907445276382,
+      "learning_rate": 9.190550124088802e-06,
+      "loss": 0.5805,
+      "step": 6961
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2395495348684942,
+      "learning_rate": 9.188045723738471e-06,
+      "loss": 0.5715,
+      "step": 6962
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0463720793240696,
+      "learning_rate": 9.185541374651001e-06,
+      "loss": 0.5656,
+      "step": 6963
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2738552227489512,
+      "learning_rate": 9.183037076984505e-06,
+      "loss": 0.5581,
+      "step": 6964
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1700438370602617,
+      "learning_rate": 9.180532830897099e-06,
+      "loss": 0.5691,
+      "step": 6965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1466987058595464,
+      "learning_rate": 9.178028636546879e-06,
+      "loss": 0.5725,
+      "step": 6966
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0806897933712363,
+      "learning_rate": 9.175524494091951e-06,
+      "loss": 0.5095,
+      "step": 6967
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.10736918113791,
+      "learning_rate": 9.173020403690417e-06,
+      "loss": 0.502,
+      "step": 6968
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1199586551562242,
+      "learning_rate": 9.17051636550037e-06,
+      "loss": 0.5631,
+      "step": 6969
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0493294141272902,
+      "learning_rate": 9.168012379679909e-06,
+      "loss": 0.5044,
+      "step": 6970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2162419457477667,
+      "learning_rate": 9.165508446387114e-06,
+      "loss": 0.5426,
+      "step": 6971
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2508817900374918,
+      "learning_rate": 9.16300456578008e-06,
+      "loss": 0.571,
+      "step": 6972
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1812503390841527,
+      "learning_rate": 9.160500738016883e-06,
+      "loss": 0.5793,
+      "step": 6973
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.274405737371036,
+      "learning_rate": 9.157996963255605e-06,
+      "loss": 0.5297,
+      "step": 6974
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.009623061839108,
+      "learning_rate": 9.155493241654327e-06,
+      "loss": 0.4643,
+      "step": 6975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2002514794214814,
+      "learning_rate": 9.152989573371115e-06,
+      "loss": 0.5486,
+      "step": 6976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2772244272290916,
+      "learning_rate": 9.150485958564043e-06,
+      "loss": 0.603,
+      "step": 6977
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1033836623426903,
+      "learning_rate": 9.147982397391172e-06,
+      "loss": 0.5454,
+      "step": 6978
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2115514367960964,
+      "learning_rate": 9.145478890010568e-06,
+      "loss": 0.571,
+      "step": 6979
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2421910987657923,
+      "learning_rate": 9.142975436580297e-06,
+      "loss": 0.5611,
+      "step": 6980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1870311514214298,
+      "learning_rate": 9.140472037258403e-06,
+      "loss": 0.5261,
+      "step": 6981
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1489866850955477,
+      "learning_rate": 9.137968692202945e-06,
+      "loss": 0.5315,
+      "step": 6982
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0719251526423996,
+      "learning_rate": 9.13546540157197e-06,
+      "loss": 0.5379,
+      "step": 6983
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3033732116107446,
+      "learning_rate": 9.132962165523523e-06,
+      "loss": 0.5357,
+      "step": 6984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.210522537503795,
+      "learning_rate": 9.130458984215652e-06,
+      "loss": 0.5308,
+      "step": 6985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2480797800612613,
+      "learning_rate": 9.12795585780639e-06,
+      "loss": 0.5878,
+      "step": 6986
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.14679730643311,
+      "learning_rate": 9.125452786453774e-06,
+      "loss": 0.5225,
+      "step": 6987
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.221749307553021,
+      "learning_rate": 9.122949770315834e-06,
+      "loss": 0.5507,
+      "step": 6988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3355400431152264,
+      "learning_rate": 9.1204468095506e-06,
+      "loss": 0.6395,
+      "step": 6989
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.282097931312729,
+      "learning_rate": 9.1179439043161e-06,
+      "loss": 0.5805,
+      "step": 6990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1581507859136013,
+      "learning_rate": 9.11544105477035e-06,
+      "loss": 0.5557,
+      "step": 6991
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2234400470263787,
+      "learning_rate": 9.11293826107137e-06,
+      "loss": 0.6168,
+      "step": 6992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1379268034170729,
+      "learning_rate": 9.110435523377173e-06,
+      "loss": 0.5406,
+      "step": 6993
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0996749614446293,
+      "learning_rate": 9.10793284184577e-06,
+      "loss": 0.5462,
+      "step": 6994
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0379292898330597,
+      "learning_rate": 9.105430216635174e-06,
+      "loss": 0.5362,
+      "step": 6995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1427434817912512,
+      "learning_rate": 9.102927647903382e-06,
+      "loss": 0.5539,
+      "step": 6996
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1408329081807391,
+      "learning_rate": 9.100425135808396e-06,
+      "loss": 0.5236,
+      "step": 6997
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0264476740935105,
+      "learning_rate": 9.09792268050821e-06,
+      "loss": 0.4896,
+      "step": 6998
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1323937003666193,
+      "learning_rate": 9.09542028216082e-06,
+      "loss": 0.487,
+      "step": 6999
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1424002745362227,
+      "learning_rate": 9.092917940924216e-06,
+      "loss": 0.5429,
+      "step": 7000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4034000764289825,
+      "learning_rate": 9.090415656956382e-06,
+      "loss": 0.5719,
+      "step": 7001
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0883032189292592,
+      "learning_rate": 9.0879134304153e-06,
+      "loss": 0.5291,
+      "step": 7002
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1988742594513302,
+      "learning_rate": 9.085411261458948e-06,
+      "loss": 0.554,
+      "step": 7003
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.155488149855366,
+      "learning_rate": 9.082909150245302e-06,
+      "loss": 0.5646,
+      "step": 7004
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2430118248822817,
+      "learning_rate": 9.080407096932334e-06,
+      "loss": 0.5543,
+      "step": 7005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.143835331386233,
+      "learning_rate": 9.077905101678011e-06,
+      "loss": 0.5131,
+      "step": 7006
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1471499017442481,
+      "learning_rate": 9.075403164640296e-06,
+      "loss": 0.5024,
+      "step": 7007
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.138680826089791,
+      "learning_rate": 9.072901285977148e-06,
+      "loss": 0.52,
+      "step": 7008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2051025585992894,
+      "learning_rate": 9.070399465846524e-06,
+      "loss": 0.5342,
+      "step": 7009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1450505038954624,
+      "learning_rate": 9.06789770440638e-06,
+      "loss": 0.5494,
+      "step": 7010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1068477588922272,
+      "learning_rate": 9.065396001814665e-06,
+      "loss": 0.5358,
+      "step": 7011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2966130003492022,
+      "learning_rate": 9.06289435822932e-06,
+      "loss": 0.6066,
+      "step": 7012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.228411889965284,
+      "learning_rate": 9.06039277380829e-06,
+      "loss": 0.5751,
+      "step": 7013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0887813517176645,
+      "learning_rate": 9.057891248709508e-06,
+      "loss": 0.5133,
+      "step": 7014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.391514911296203,
+      "learning_rate": 9.055389783090916e-06,
+      "loss": 0.5703,
+      "step": 7015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1722422723627295,
+      "learning_rate": 9.05288837711044e-06,
+      "loss": 0.532,
+      "step": 7016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.085709692023892,
+      "learning_rate": 9.050387030926008e-06,
+      "loss": 0.5014,
+      "step": 7017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1962447177686633,
+      "learning_rate": 9.047885744695539e-06,
+      "loss": 0.5419,
+      "step": 7018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1480357706597895,
+      "learning_rate": 9.045384518576956e-06,
+      "loss": 0.5024,
+      "step": 7019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.20420491686902,
+      "learning_rate": 9.042883352728176e-06,
+      "loss": 0.5586,
+      "step": 7020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1029276944252733,
+      "learning_rate": 9.040382247307107e-06,
+      "loss": 0.5225,
+      "step": 7021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.311248409608471,
+      "learning_rate": 9.037881202471657e-06,
+      "loss": 0.6048,
+      "step": 7022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1548144733989967,
+      "learning_rate": 9.03538021837973e-06,
+      "loss": 0.5379,
+      "step": 7023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1492036612302654,
+      "learning_rate": 9.032879295189226e-06,
+      "loss": 0.5451,
+      "step": 7024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2538196854702794,
+      "learning_rate": 9.030378433058045e-06,
+      "loss": 0.6014,
+      "step": 7025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1552907212061672,
+      "learning_rate": 9.027877632144076e-06,
+      "loss": 0.5065,
+      "step": 7026
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1356965282967875,
+      "learning_rate": 9.025376892605205e-06,
+      "loss": 0.5296,
+      "step": 7027
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2273073303918798,
+      "learning_rate": 9.02287621459932e-06,
+      "loss": 0.5217,
+      "step": 7028
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1045813351649905,
+      "learning_rate": 9.0203755982843e-06,
+      "loss": 0.5463,
+      "step": 7029
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2339777729839565,
+      "learning_rate": 9.017875043818024e-06,
+      "loss": 0.5377,
+      "step": 7030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0785493982892813,
+      "learning_rate": 9.015374551358364e-06,
+      "loss": 0.5226,
+      "step": 7031
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1916739741459232,
+      "learning_rate": 9.012874121063189e-06,
+      "loss": 0.5848,
+      "step": 7032
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.204853896389094,
+      "learning_rate": 9.01037375309036e-06,
+      "loss": 0.5662,
+      "step": 7033
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2784209302874183,
+      "learning_rate": 9.007873447597744e-06,
+      "loss": 0.5707,
+      "step": 7034
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.143587107080752,
+      "learning_rate": 9.005373204743196e-06,
+      "loss": 0.5527,
+      "step": 7035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0765289987962836,
+      "learning_rate": 9.00287302468457e-06,
+      "loss": 0.4727,
+      "step": 7036
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1779686234607318,
+      "learning_rate": 9.000372907579716e-06,
+      "loss": 0.5093,
+      "step": 7037
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.199891687511814,
+      "learning_rate": 8.997872853586474e-06,
+      "loss": 0.5887,
+      "step": 7038
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1354772860708937,
+      "learning_rate": 8.995372862862687e-06,
+      "loss": 0.4775,
+      "step": 7039
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1514607812892144,
+      "learning_rate": 8.9928729355662e-06,
+      "loss": 0.5055,
+      "step": 7040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2774361393264921,
+      "learning_rate": 8.990373071854842e-06,
+      "loss": 0.6212,
+      "step": 7041
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2539147112314308,
+      "learning_rate": 8.987873271886436e-06,
+      "loss": 0.567,
+      "step": 7042
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0432130334641108,
+      "learning_rate": 8.985373535818814e-06,
+      "loss": 0.5036,
+      "step": 7043
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2721665950248542,
+      "learning_rate": 8.982873863809793e-06,
+      "loss": 0.5396,
+      "step": 7044
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1076160509604467,
+      "learning_rate": 8.980374256017196e-06,
+      "loss": 0.4962,
+      "step": 7045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2018725824753307,
+      "learning_rate": 8.977874712598833e-06,
+      "loss": 0.5538,
+      "step": 7046
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.137887360567805,
+      "learning_rate": 8.975375233712511e-06,
+      "loss": 0.5593,
+      "step": 7047
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1979539395490664,
+      "learning_rate": 8.972875819516037e-06,
+      "loss": 0.5256,
+      "step": 7048
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.271288269202131,
+      "learning_rate": 8.97037647016721e-06,
+      "loss": 0.56,
+      "step": 7049
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1684079062261288,
+      "learning_rate": 8.967877185823833e-06,
+      "loss": 0.5269,
+      "step": 7050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1487415975070363,
+      "learning_rate": 8.96537796664369e-06,
+      "loss": 0.5491,
+      "step": 7051
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0950324578025312,
+      "learning_rate": 8.96287881278458e-06,
+      "loss": 0.5636,
+      "step": 7052
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1393221377415492,
+      "learning_rate": 8.960379724404275e-06,
+      "loss": 0.5054,
+      "step": 7053
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2108883755471758,
+      "learning_rate": 8.957880701660563e-06,
+      "loss": 0.5465,
+      "step": 7054
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2598750577015054,
+      "learning_rate": 8.955381744711222e-06,
+      "loss": 0.5956,
+      "step": 7055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2598406155762636,
+      "learning_rate": 8.952882853714017e-06,
+      "loss": 0.5896,
+      "step": 7056
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1970730948793176,
+      "learning_rate": 8.950384028826724e-06,
+      "loss": 0.56,
+      "step": 7057
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.226250432643367,
+      "learning_rate": 8.9478852702071e-06,
+      "loss": 0.6067,
+      "step": 7058
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1200959553656014,
+      "learning_rate": 8.945386578012906e-06,
+      "loss": 0.532,
+      "step": 7059
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0510084622711193,
+      "learning_rate": 8.942887952401901e-06,
+      "loss": 0.4692,
+      "step": 7060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1454068632921695,
+      "learning_rate": 8.94038939353183e-06,
+      "loss": 0.5528,
+      "step": 7061
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1907246124969835,
+      "learning_rate": 8.937890901560452e-06,
+      "loss": 0.5423,
+      "step": 7062
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2802901045051178,
+      "learning_rate": 8.935392476645493e-06,
+      "loss": 0.5969,
+      "step": 7063
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1649678952782836,
+      "learning_rate": 8.932894118944699e-06,
+      "loss": 0.5564,
+      "step": 7064
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1245136269252156,
+      "learning_rate": 8.930395828615808e-06,
+      "loss": 0.5204,
+      "step": 7065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.158987413382461,
+      "learning_rate": 8.927897605816546e-06,
+      "loss": 0.5468,
+      "step": 7066
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1992520007764307,
+      "learning_rate": 8.925399450704642e-06,
+      "loss": 0.5747,
+      "step": 7067
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0992629486286078,
+      "learning_rate": 8.922901363437812e-06,
+      "loss": 0.5121,
+      "step": 7068
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.087809205218736,
+      "learning_rate": 8.920403344173776e-06,
+      "loss": 0.5131,
+      "step": 7069
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.15644850186857,
+      "learning_rate": 8.917905393070249e-06,
+      "loss": 0.4964,
+      "step": 7070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1522958519242956,
+      "learning_rate": 8.915407510284933e-06,
+      "loss": 0.4904,
+      "step": 7071
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0675906915130946,
+      "learning_rate": 8.912909695975543e-06,
+      "loss": 0.561,
+      "step": 7072
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1408533886655317,
+      "learning_rate": 8.910411950299769e-06,
+      "loss": 0.4897,
+      "step": 7073
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0958324770809256,
+      "learning_rate": 8.907914273415308e-06,
+      "loss": 0.527,
+      "step": 7074
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2821511146910982,
+      "learning_rate": 8.905416665479856e-06,
+      "loss": 0.5092,
+      "step": 7075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2290909091858149,
+      "learning_rate": 8.902919126651096e-06,
+      "loss": 0.5701,
+      "step": 7076
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2122036129224039,
+      "learning_rate": 8.900421657086716e-06,
+      "loss": 0.5234,
+      "step": 7077
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1998593983137111,
+      "learning_rate": 8.897924256944386e-06,
+      "loss": 0.5974,
+      "step": 7078
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1215766637571554,
+      "learning_rate": 8.895426926381782e-06,
+      "loss": 0.5151,
+      "step": 7079
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.086803447787699,
+      "learning_rate": 8.892929665556577e-06,
+      "loss": 0.5216,
+      "step": 7080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0911354968811233,
+      "learning_rate": 8.890432474626433e-06,
+      "loss": 0.4995,
+      "step": 7081
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1227341722431547,
+      "learning_rate": 8.887935353749017e-06,
+      "loss": 0.5727,
+      "step": 7082
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3099574939733143,
+      "learning_rate": 8.885438303081972e-06,
+      "loss": 0.5852,
+      "step": 7083
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2274453454368328,
+      "learning_rate": 8.88294132278296e-06,
+      "loss": 0.5978,
+      "step": 7084
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1841064450608139,
+      "learning_rate": 8.880444413009627e-06,
+      "loss": 0.5602,
+      "step": 7085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3034456933330383,
+      "learning_rate": 8.877947573919612e-06,
+      "loss": 0.5742,
+      "step": 7086
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2213081020406638,
+      "learning_rate": 8.87545080567056e-06,
+      "loss": 0.5324,
+      "step": 7087
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2171063221022014,
+      "learning_rate": 8.872954108420096e-06,
+      "loss": 0.5474,
+      "step": 7088
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0927276738106495,
+      "learning_rate": 8.870457482325854e-06,
+      "loss": 0.534,
+      "step": 7089
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1417210159916185,
+      "learning_rate": 8.867960927545461e-06,
+      "loss": 0.52,
+      "step": 7090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1728181731007823,
+      "learning_rate": 8.865464444236534e-06,
+      "loss": 0.5447,
+      "step": 7091
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.076641554898837,
+      "learning_rate": 8.862968032556694e-06,
+      "loss": 0.4897,
+      "step": 7092
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1660198166866136,
+      "learning_rate": 8.860471692663542e-06,
+      "loss": 0.5194,
+      "step": 7093
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1576782375229382,
+      "learning_rate": 8.857975424714694e-06,
+      "loss": 0.5384,
+      "step": 7094
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2249639019221212,
+      "learning_rate": 8.855479228867751e-06,
+      "loss": 0.5313,
+      "step": 7095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.208790999371075,
+      "learning_rate": 8.852983105280307e-06,
+      "loss": 0.5283,
+      "step": 7096
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0993255194653686,
+      "learning_rate": 8.850487054109961e-06,
+      "loss": 0.5268,
+      "step": 7097
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1716331232314943,
+      "learning_rate": 8.847991075514294e-06,
+      "loss": 0.5317,
+      "step": 7098
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1810950162426344,
+      "learning_rate": 8.845495169650894e-06,
+      "loss": 0.5561,
+      "step": 7099
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.268011786011996,
+      "learning_rate": 8.842999336677342e-06,
+      "loss": 0.591,
+      "step": 7100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.239070171702857,
+      "learning_rate": 8.84050357675121e-06,
+      "loss": 0.5664,
+      "step": 7101
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2000389152415756,
+      "learning_rate": 8.838007890030074e-06,
+      "loss": 0.5522,
+      "step": 7102
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1787176124111955,
+      "learning_rate": 8.83551227667149e-06,
+      "loss": 0.5402,
+      "step": 7103
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2485091856044537,
+      "learning_rate": 8.833016736833023e-06,
+      "loss": 0.5624,
+      "step": 7104
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.257974507578442,
+      "learning_rate": 8.830521270672233e-06,
+      "loss": 0.5433,
+      "step": 7105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1308969020381792,
+      "learning_rate": 8.828025878346667e-06,
+      "loss": 0.5292,
+      "step": 7106
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1160898218001667,
+      "learning_rate": 8.825530560013876e-06,
+      "loss": 0.4983,
+      "step": 7107
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1349720857355323,
+      "learning_rate": 8.8230353158314e-06,
+      "loss": 0.578,
+      "step": 7108
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.091982475978547,
+      "learning_rate": 8.820540145956776e-06,
+      "loss": 0.465,
+      "step": 7109
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2091692904797267,
+      "learning_rate": 8.818045050547539e-06,
+      "loss": 0.5271,
+      "step": 7110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1360981613784953,
+      "learning_rate": 8.815550029761215e-06,
+      "loss": 0.5331,
+      "step": 7111
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.138158090142865,
+      "learning_rate": 8.813055083755327e-06,
+      "loss": 0.5017,
+      "step": 7112
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1919370875717665,
+      "learning_rate": 8.8105602126874e-06,
+      "loss": 0.5528,
+      "step": 7113
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.064462924177229,
+      "learning_rate": 8.808065416714938e-06,
+      "loss": 0.4861,
+      "step": 7114
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1121106334340165,
+      "learning_rate": 8.805570695995462e-06,
+      "loss": 0.4776,
+      "step": 7115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1453971841893755,
+      "learning_rate": 8.803076050686465e-06,
+      "loss": 0.542,
+      "step": 7116
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1840272116014923,
+      "learning_rate": 8.800581480945453e-06,
+      "loss": 0.5305,
+      "step": 7117
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3058131952361969,
+      "learning_rate": 8.798086986929923e-06,
+      "loss": 0.5811,
+      "step": 7118
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1610344539527588,
+      "learning_rate": 8.795592568797362e-06,
+      "loss": 0.5349,
+      "step": 7119
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1838873571444342,
+      "learning_rate": 8.793098226705255e-06,
+      "loss": 0.5324,
+      "step": 7120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2025679252299515,
+      "learning_rate": 8.790603960811082e-06,
+      "loss": 0.5428,
+      "step": 7121
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1145408063719777,
+      "learning_rate": 8.78810977127232e-06,
+      "loss": 0.5201,
+      "step": 7122
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1309701078024728,
+      "learning_rate": 8.785615658246448e-06,
+      "loss": 0.5456,
+      "step": 7123
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1486368330710166,
+      "learning_rate": 8.783121621890917e-06,
+      "loss": 0.5575,
+      "step": 7124
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.256711584666228,
+      "learning_rate": 8.780627662363201e-06,
+      "loss": 0.58,
+      "step": 7125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1477685134647573,
+      "learning_rate": 8.77813377982075e-06,
+      "loss": 0.5745,
+      "step": 7126
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1382409355025513,
+      "learning_rate": 8.775639974421015e-06,
+      "loss": 0.596,
+      "step": 7127
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1452729617758346,
+      "learning_rate": 8.773146246321452e-06,
+      "loss": 0.5872,
+      "step": 7128
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1428474304518808,
+      "learning_rate": 8.77065259567949e-06,
+      "loss": 0.547,
+      "step": 7129
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.28632688687355,
+      "learning_rate": 8.768159022652574e-06,
+      "loss": 0.6348,
+      "step": 7130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2055499422193672,
+      "learning_rate": 8.765665527398133e-06,
+      "loss": 0.5535,
+      "step": 7131
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0953780455302633,
+      "learning_rate": 8.763172110073596e-06,
+      "loss": 0.5218,
+      "step": 7132
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0788903215186887,
+      "learning_rate": 8.760678770836391e-06,
+      "loss": 0.519,
+      "step": 7133
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1224003000310658,
+      "learning_rate": 8.758185509843921e-06,
+      "loss": 0.4871,
+      "step": 7134
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0695437155443668,
+      "learning_rate": 8.755692327253612e-06,
+      "loss": 0.5097,
+      "step": 7135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1795728008770674,
+      "learning_rate": 8.753199223222863e-06,
+      "loss": 0.5573,
+      "step": 7136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2191648510814888,
+      "learning_rate": 8.75070619790908e-06,
+      "loss": 0.5931,
+      "step": 7137
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1634564454172729,
+      "learning_rate": 8.748213251469663e-06,
+      "loss": 0.5397,
+      "step": 7138
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1976217266165836,
+      "learning_rate": 8.745720384062002e-06,
+      "loss": 0.5591,
+      "step": 7139
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.046246183199284,
+      "learning_rate": 8.74322759584348e-06,
+      "loss": 0.4605,
+      "step": 7140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1382486855817477,
+      "learning_rate": 8.740734886971485e-06,
+      "loss": 0.528,
+      "step": 7141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1312567189053966,
+      "learning_rate": 8.738242257603394e-06,
+      "loss": 0.5534,
+      "step": 7142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0714343797435453,
+      "learning_rate": 8.735749707896584e-06,
+      "loss": 0.4893,
+      "step": 7143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.27354393262686,
+      "learning_rate": 8.733257238008414e-06,
+      "loss": 0.6157,
+      "step": 7144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.151990519550838,
+      "learning_rate": 8.730764848096247e-06,
+      "loss": 0.5211,
+      "step": 7145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1484824450774165,
+      "learning_rate": 8.728272538317447e-06,
+      "loss": 0.5343,
+      "step": 7146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2753813248159431,
+      "learning_rate": 8.725780308829358e-06,
+      "loss": 0.5678,
+      "step": 7147
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0761916475504654,
+      "learning_rate": 8.72328815978934e-06,
+      "loss": 0.5449,
+      "step": 7148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2135464036491619,
+      "learning_rate": 8.720796091354725e-06,
+      "loss": 0.536,
+      "step": 7149
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.290066830656032,
+      "learning_rate": 8.71830410368285e-06,
+      "loss": 0.5861,
+      "step": 7150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1461821747559466,
+      "learning_rate": 8.715812196931049e-06,
+      "loss": 0.5073,
+      "step": 7151
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1372245371197143,
+      "learning_rate": 8.713320371256651e-06,
+      "loss": 0.5703,
+      "step": 7152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.227331661405224,
+      "learning_rate": 8.71082862681698e-06,
+      "loss": 0.5832,
+      "step": 7153
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2050269810042098,
+      "learning_rate": 8.708336963769345e-06,
+      "loss": 0.5622,
+      "step": 7154
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1607089283578502,
+      "learning_rate": 8.705845382271063e-06,
+      "loss": 0.5249,
+      "step": 7155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2006896500902056,
+      "learning_rate": 8.703353882479436e-06,
+      "loss": 0.6039,
+      "step": 7156
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1997359958951372,
+      "learning_rate": 8.70086246455177e-06,
+      "loss": 0.523,
+      "step": 7157
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.149052876755601,
+      "learning_rate": 8.698371128645364e-06,
+      "loss": 0.57,
+      "step": 7158
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2437832257610775,
+      "learning_rate": 8.6958798749175e-06,
+      "loss": 0.5645,
+      "step": 7159
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2376319314007216,
+      "learning_rate": 8.693388703525467e-06,
+      "loss": 0.5769,
+      "step": 7160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1158212163284837,
+      "learning_rate": 8.690897614626546e-06,
+      "loss": 0.5515,
+      "step": 7161
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3077068454810339,
+      "learning_rate": 8.688406608378012e-06,
+      "loss": 0.5694,
+      "step": 7162
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1859861562445195,
+      "learning_rate": 8.685915684937138e-06,
+      "loss": 0.5276,
+      "step": 7163
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1966382322855935,
+      "learning_rate": 8.68342484446119e-06,
+      "loss": 0.5182,
+      "step": 7164
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1723239293101348,
+      "learning_rate": 8.68093408710742e-06,
+      "loss": 0.5786,
+      "step": 7165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2801214225368167,
+      "learning_rate": 8.678443413033085e-06,
+      "loss": 0.5472,
+      "step": 7166
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1055663911542983,
+      "learning_rate": 8.675952822395437e-06,
+      "loss": 0.4865,
+      "step": 7167
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2392195266569208,
+      "learning_rate": 8.67346231535172e-06,
+      "loss": 0.5382,
+      "step": 7168
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1553552103647624,
+      "learning_rate": 8.670971892059173e-06,
+      "loss": 0.5446,
+      "step": 7169
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0555646286680438,
+      "learning_rate": 8.668481552675024e-06,
+      "loss": 0.4894,
+      "step": 7170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.169164906258231,
+      "learning_rate": 8.665991297356503e-06,
+      "loss": 0.5314,
+      "step": 7171
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2251974394366452,
+      "learning_rate": 8.663501126260836e-06,
+      "loss": 0.5657,
+      "step": 7172
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0571001502656456,
+      "learning_rate": 8.661011039545238e-06,
+      "loss": 0.5271,
+      "step": 7173
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1893826668953762,
+      "learning_rate": 8.658521037366926e-06,
+      "loss": 0.5124,
+      "step": 7174
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0664217322486766,
+      "learning_rate": 8.656031119883095e-06,
+      "loss": 0.5068,
+      "step": 7175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2050477059604079,
+      "learning_rate": 8.653541287250954e-06,
+      "loss": 0.5621,
+      "step": 7176
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0894080490903262,
+      "learning_rate": 8.6510515396277e-06,
+      "loss": 0.4987,
+      "step": 7177
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0904926296231676,
+      "learning_rate": 8.648561877170522e-06,
+      "loss": 0.4856,
+      "step": 7178
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0408307281677367,
+      "learning_rate": 8.64607230003661e-06,
+      "loss": 0.4872,
+      "step": 7179
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1827934757451357,
+      "learning_rate": 8.643582808383133e-06,
+      "loss": 0.5246,
+      "step": 7180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3385979112418176,
+      "learning_rate": 8.641093402367272e-06,
+      "loss": 0.55,
+      "step": 7181
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1946439681128436,
+      "learning_rate": 8.638604082146195e-06,
+      "loss": 0.6115,
+      "step": 7182
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2237109915075493,
+      "learning_rate": 8.636114847877068e-06,
+      "loss": 0.5065,
+      "step": 7183
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3110380204664942,
+      "learning_rate": 8.633625699717051e-06,
+      "loss": 0.5248,
+      "step": 7184
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0426367756428179,
+      "learning_rate": 8.631136637823288e-06,
+      "loss": 0.5071,
+      "step": 7185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1508867783857837,
+      "learning_rate": 8.628647662352932e-06,
+      "loss": 0.5257,
+      "step": 7186
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1011325186374112,
+      "learning_rate": 8.626158773463124e-06,
+      "loss": 0.539,
+      "step": 7187
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1338877869041564,
+      "learning_rate": 8.623669971311002e-06,
+      "loss": 0.4981,
+      "step": 7188
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2576198548823194,
+      "learning_rate": 8.621181256053699e-06,
+      "loss": 0.5641,
+      "step": 7189
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.163963057603609,
+      "learning_rate": 8.618692627848331e-06,
+      "loss": 0.5258,
+      "step": 7190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2466958722775159,
+      "learning_rate": 8.616204086852026e-06,
+      "loss": 0.5606,
+      "step": 7191
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2311160368099745,
+      "learning_rate": 8.613715633221895e-06,
+      "loss": 0.5969,
+      "step": 7192
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1809099949351327,
+      "learning_rate": 8.611227267115052e-06,
+      "loss": 0.5437,
+      "step": 7193
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2797272169496263,
+      "learning_rate": 8.608738988688598e-06,
+      "loss": 0.576,
+      "step": 7194
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1589875162388705,
+      "learning_rate": 8.606250798099626e-06,
+      "loss": 0.5715,
+      "step": 7195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2012791531426092,
+      "learning_rate": 8.603762695505231e-06,
+      "loss": 0.5358,
+      "step": 7196
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1834082631220921,
+      "learning_rate": 8.601274681062502e-06,
+      "loss": 0.5103,
+      "step": 7197
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.204446092060373,
+      "learning_rate": 8.598786754928519e-06,
+      "loss": 0.4844,
+      "step": 7198
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1348855354130147,
+      "learning_rate": 8.596298917260361e-06,
+      "loss": 0.5131,
+      "step": 7199
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2509721313683964,
+      "learning_rate": 8.59381116821509e-06,
+      "loss": 0.5928,
+      "step": 7200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2113007523618666,
+      "learning_rate": 8.591323507949773e-06,
+      "loss": 0.5475,
+      "step": 7201
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.113959764877169,
+      "learning_rate": 8.588835936621473e-06,
+      "loss": 0.5209,
+      "step": 7202
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.149132861902752,
+      "learning_rate": 8.586348454387244e-06,
+      "loss": 0.5237,
+      "step": 7203
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1629394131473179,
+      "learning_rate": 8.583861061404131e-06,
+      "loss": 0.552,
+      "step": 7204
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0557785042725218,
+      "learning_rate": 8.581373757829172e-06,
+      "loss": 0.5328,
+      "step": 7205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2137346509911722,
+      "learning_rate": 8.578886543819406e-06,
+      "loss": 0.5114,
+      "step": 7206
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0513011114370965,
+      "learning_rate": 8.576399419531865e-06,
+      "loss": 0.4742,
+      "step": 7207
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2870882579593583,
+      "learning_rate": 8.573912385123576e-06,
+      "loss": 0.5477,
+      "step": 7208
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2033791707062595,
+      "learning_rate": 8.571425440751557e-06,
+      "loss": 0.5532,
+      "step": 7209
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1112153275941123,
+      "learning_rate": 8.568938586572816e-06,
+      "loss": 0.544,
+      "step": 7210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.269166113355937,
+      "learning_rate": 8.566451822744367e-06,
+      "loss": 0.5632,
+      "step": 7211
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.103377341999504,
+      "learning_rate": 8.563965149423207e-06,
+      "loss": 0.5588,
+      "step": 7212
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1829084168500636,
+      "learning_rate": 8.561478566766341e-06,
+      "loss": 0.5701,
+      "step": 7213
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1282394182083846,
+      "learning_rate": 8.558992074930757e-06,
+      "loss": 0.555,
+      "step": 7214
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1897374206442264,
+      "learning_rate": 8.556505674073435e-06,
+      "loss": 0.5654,
+      "step": 7215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1373050396863185,
+      "learning_rate": 8.554019364351354e-06,
+      "loss": 0.5329,
+      "step": 7216
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2656010284037458,
+      "learning_rate": 8.551533145921493e-06,
+      "loss": 0.5823,
+      "step": 7217
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0771660894087123,
+      "learning_rate": 8.54904701894082e-06,
+      "loss": 0.5265,
+      "step": 7218
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1263101154842148,
+      "learning_rate": 8.54656098356629e-06,
+      "loss": 0.5315,
+      "step": 7219
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.190866566980221,
+      "learning_rate": 8.54407503995487e-06,
+      "loss": 0.6044,
+      "step": 7220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1645208390903092,
+      "learning_rate": 8.541589188263499e-06,
+      "loss": 0.5599,
+      "step": 7221
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3026463880665522,
+      "learning_rate": 8.53910342864913e-06,
+      "loss": 0.5796,
+      "step": 7222
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1717749998023268,
+      "learning_rate": 8.536617761268697e-06,
+      "loss": 0.5296,
+      "step": 7223
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1187839438973723,
+      "learning_rate": 8.534132186279134e-06,
+      "loss": 0.4973,
+      "step": 7224
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1452520919211118,
+      "learning_rate": 8.531646703837375e-06,
+      "loss": 0.515,
+      "step": 7225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.149693120562664,
+      "learning_rate": 8.52916131410033e-06,
+      "loss": 0.5897,
+      "step": 7226
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2401958790008578,
+      "learning_rate": 8.526676017224917e-06,
+      "loss": 0.4929,
+      "step": 7227
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0882690976704232,
+      "learning_rate": 8.524190813368055e-06,
+      "loss": 0.5132,
+      "step": 7228
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1196918295186495,
+      "learning_rate": 8.521705702686636e-06,
+      "loss": 0.5099,
+      "step": 7229
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0947322794265588,
+      "learning_rate": 8.519220685337567e-06,
+      "loss": 0.4871,
+      "step": 7230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1736105718868504,
+      "learning_rate": 8.516735761477734e-06,
+      "loss": 0.5403,
+      "step": 7231
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1571712432386958,
+      "learning_rate": 8.514250931264023e-06,
+      "loss": 0.5147,
+      "step": 7232
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.207567213253128,
+      "learning_rate": 8.51176619485332e-06,
+      "loss": 0.5414,
+      "step": 7233
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1109181806169475,
+      "learning_rate": 8.509281552402492e-06,
+      "loss": 0.4779,
+      "step": 7234
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1273065868869139,
+      "learning_rate": 8.506797004068414e-06,
+      "loss": 0.5445,
+      "step": 7235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1870942175381705,
+      "learning_rate": 8.504312550007943e-06,
+      "loss": 0.5392,
+      "step": 7236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.144063964859484,
+      "learning_rate": 8.501828190377936e-06,
+      "loss": 0.5915,
+      "step": 7237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2614229877202074,
+      "learning_rate": 8.499343925335249e-06,
+      "loss": 0.5357,
+      "step": 7238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0709019729246994,
+      "learning_rate": 8.496859755036719e-06,
+      "loss": 0.5138,
+      "step": 7239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2774763126006197,
+      "learning_rate": 8.49437567963919e-06,
+      "loss": 0.5745,
+      "step": 7240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0626795560919822,
+      "learning_rate": 8.491891699299491e-06,
+      "loss": 0.5294,
+      "step": 7241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0893981460152122,
+      "learning_rate": 8.48940781417445e-06,
+      "loss": 0.5322,
+      "step": 7242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.17873657502087,
+      "learning_rate": 8.486924024420887e-06,
+      "loss": 0.5115,
+      "step": 7243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1926663098435129,
+      "learning_rate": 8.484440330195615e-06,
+      "loss": 0.528,
+      "step": 7244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1950658406590542,
+      "learning_rate": 8.481956731655451e-06,
+      "loss": 0.4774,
+      "step": 7245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1552145677724501,
+      "learning_rate": 8.479473228957185e-06,
+      "loss": 0.5308,
+      "step": 7246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1907529947362159,
+      "learning_rate": 8.47698982225762e-06,
+      "loss": 0.5542,
+      "step": 7247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1525258065931892,
+      "learning_rate": 8.474506511713543e-06,
+      "loss": 0.5582,
+      "step": 7248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2058761155597932,
+      "learning_rate": 8.472023297481741e-06,
+      "loss": 0.5131,
+      "step": 7249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1127829524045973,
+      "learning_rate": 8.469540179718997e-06,
+      "loss": 0.5099,
+      "step": 7250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2010304933785343,
+      "learning_rate": 8.467057158582072e-06,
+      "loss": 0.5843,
+      "step": 7251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3161147612735844,
+      "learning_rate": 8.46457423422774e-06,
+      "loss": 0.5905,
+      "step": 7252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.128347872542259,
+      "learning_rate": 8.462091406812759e-06,
+      "loss": 0.5235,
+      "step": 7253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0829132561644017,
+      "learning_rate": 8.459608676493878e-06,
+      "loss": 0.4717,
+      "step": 7254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1453541476183215,
+      "learning_rate": 8.457126043427855e-06,
+      "loss": 0.5432,
+      "step": 7255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2237496164260901,
+      "learning_rate": 8.45464350777142e-06,
+      "loss": 0.5754,
+      "step": 7256
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1572886776248892,
+      "learning_rate": 8.452161069681315e-06,
+      "loss": 0.5131,
+      "step": 7257
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2030709985280958,
+      "learning_rate": 8.449678729314266e-06,
+      "loss": 0.5472,
+      "step": 7258
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1487454371315013,
+      "learning_rate": 8.447196486826996e-06,
+      "loss": 0.533,
+      "step": 7259
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0968306073283058,
+      "learning_rate": 8.44471434237623e-06,
+      "loss": 0.4957,
+      "step": 7260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.175836099251446,
+      "learning_rate": 8.442232296118667e-06,
+      "loss": 0.5089,
+      "step": 7261
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1631159678441056,
+      "learning_rate": 8.439750348211016e-06,
+      "loss": 0.5248,
+      "step": 7262
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.183753377391988,
+      "learning_rate": 8.437268498809975e-06,
+      "loss": 0.5329,
+      "step": 7263
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2399476688632751,
+      "learning_rate": 8.434786748072237e-06,
+      "loss": 0.6115,
+      "step": 7264
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0345711836073948,
+      "learning_rate": 8.43230509615449e-06,
+      "loss": 0.5,
+      "step": 7265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2119532601090794,
+      "learning_rate": 8.429823543213406e-06,
+      "loss": 0.5383,
+      "step": 7266
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0496316172745663,
+      "learning_rate": 8.427342089405667e-06,
+      "loss": 0.4923,
+      "step": 7267
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.192263985476433,
+      "learning_rate": 8.424860734887932e-06,
+      "loss": 0.5526,
+      "step": 7268
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2567423657469419,
+      "learning_rate": 8.422379479816865e-06,
+      "loss": 0.5885,
+      "step": 7269
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2610462865025873,
+      "learning_rate": 8.419898324349122e-06,
+      "loss": 0.492,
+      "step": 7270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1361697729511353,
+      "learning_rate": 8.417417268641354e-06,
+      "loss": 0.5552,
+      "step": 7271
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1669624101359166,
+      "learning_rate": 8.414936312850196e-06,
+      "loss": 0.5574,
+      "step": 7272
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2420983432467272,
+      "learning_rate": 8.412455457132285e-06,
+      "loss": 0.5897,
+      "step": 7273
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2690424990680642,
+      "learning_rate": 8.409974701644251e-06,
+      "loss": 0.5629,
+      "step": 7274
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.097771146469405,
+      "learning_rate": 8.40749404654272e-06,
+      "loss": 0.5428,
+      "step": 7275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1377758865311753,
+      "learning_rate": 8.405013491984307e-06,
+      "loss": 0.5241,
+      "step": 7276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2046186909895835,
+      "learning_rate": 8.40253303812562e-06,
+      "loss": 0.5155,
+      "step": 7277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.18020295256656,
+      "learning_rate": 8.400052685123263e-06,
+      "loss": 0.5699,
+      "step": 7278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1504131632282575,
+      "learning_rate": 8.397572433133836e-06,
+      "loss": 0.5481,
+      "step": 7279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2071228980174176,
+      "learning_rate": 8.395092282313927e-06,
+      "loss": 0.5788,
+      "step": 7280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0344017878172516,
+      "learning_rate": 8.392612232820125e-06,
+      "loss": 0.4855,
+      "step": 7281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1339002976899328,
+      "learning_rate": 8.390132284809005e-06,
+      "loss": 0.538,
+      "step": 7282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1394780807194689,
+      "learning_rate": 8.387652438437138e-06,
+      "loss": 0.536,
+      "step": 7283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.139242196266246,
+      "learning_rate": 8.385172693861092e-06,
+      "loss": 0.5599,
+      "step": 7284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1531086080569763,
+      "learning_rate": 8.382693051237424e-06,
+      "loss": 0.5404,
+      "step": 7285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1701272774612959,
+      "learning_rate": 8.380213510722692e-06,
+      "loss": 0.5402,
+      "step": 7286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3134095809477861,
+      "learning_rate": 8.377734072473437e-06,
+      "loss": 0.6476,
+      "step": 7287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1960375213861296,
+      "learning_rate": 8.375254736646197e-06,
+      "loss": 0.5625,
+      "step": 7288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.157223678134591,
+      "learning_rate": 8.372775503397507e-06,
+      "loss": 0.5332,
+      "step": 7289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0540106402842708,
+      "learning_rate": 8.370296372883898e-06,
+      "loss": 0.5107,
+      "step": 7290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1678402424927279,
+      "learning_rate": 8.367817345261888e-06,
+      "loss": 0.5454,
+      "step": 7291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2041928890149463,
+      "learning_rate": 8.36533842068799e-06,
+      "loss": 0.5472,
+      "step": 7292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1474760533412618,
+      "learning_rate": 8.362859599318708e-06,
+      "loss": 0.5468,
+      "step": 7293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2105314496889517,
+      "learning_rate": 8.360380881310545e-06,
+      "loss": 0.5101,
+      "step": 7294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.192729827598338,
+      "learning_rate": 8.357902266819999e-06,
+      "loss": 0.5766,
+      "step": 7295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3202619373639282,
+      "learning_rate": 8.355423756003557e-06,
+      "loss": 0.5666,
+      "step": 7296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.123303777087027,
+      "learning_rate": 8.352945349017699e-06,
+      "loss": 0.5174,
+      "step": 7297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0945743179288843,
+      "learning_rate": 8.350467046018892e-06,
+      "loss": 0.5087,
+      "step": 7298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1474797933124947,
+      "learning_rate": 8.347988847163615e-06,
+      "loss": 0.5414,
+      "step": 7299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0913709652548347,
+      "learning_rate": 8.345510752608324e-06,
+      "loss": 0.503,
+      "step": 7300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1080336722549131,
+      "learning_rate": 8.343032762509478e-06,
+      "loss": 0.5209,
+      "step": 7301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.13434097827514,
+      "learning_rate": 8.340554877023523e-06,
+      "loss": 0.5619,
+      "step": 7302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0580863722066831,
+      "learning_rate": 8.338077096306895e-06,
+      "loss": 0.4432,
+      "step": 7303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.112115242677624,
+      "learning_rate": 8.335599420516036e-06,
+      "loss": 0.5087,
+      "step": 7304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1741007711972185,
+      "learning_rate": 8.333121849807374e-06,
+      "loss": 0.5509,
+      "step": 7305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.229404582689245,
+      "learning_rate": 8.330644384337334e-06,
+      "loss": 0.5892,
+      "step": 7306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.163419353826195,
+      "learning_rate": 8.328167024262323e-06,
+      "loss": 0.5503,
+      "step": 7307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2644014448718197,
+      "learning_rate": 8.32568976973875e-06,
+      "loss": 0.5323,
+      "step": 7308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.204033595237437,
+      "learning_rate": 8.323212620923023e-06,
+      "loss": 0.5449,
+      "step": 7309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1461771824794256,
+      "learning_rate": 8.320735577971533e-06,
+      "loss": 0.5388,
+      "step": 7310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2193195772700351,
+      "learning_rate": 8.318258641040674e-06,
+      "loss": 0.5264,
+      "step": 7311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2366653633282179,
+      "learning_rate": 8.31578181028682e-06,
+      "loss": 0.5255,
+      "step": 7312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0600838396816714,
+      "learning_rate": 8.31330508586635e-06,
+      "loss": 0.5186,
+      "step": 7313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3015407143700979,
+      "learning_rate": 8.31082846793563e-06,
+      "loss": 0.6232,
+      "step": 7314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.07671999948839,
+      "learning_rate": 8.308351956651026e-06,
+      "loss": 0.5044,
+      "step": 7315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5728134811947052,
+      "learning_rate": 8.305875552168894e-06,
+      "loss": 0.5224,
+      "step": 7316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1416076190092765,
+      "learning_rate": 8.303399254645578e-06,
+      "loss": 0.4906,
+      "step": 7317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1407239688107764,
+      "learning_rate": 8.300923064237417e-06,
+      "loss": 0.5019,
+      "step": 7318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2180392809517295,
+      "learning_rate": 8.298446981100749e-06,
+      "loss": 0.5711,
+      "step": 7319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1200144287440175,
+      "learning_rate": 8.295971005391902e-06,
+      "loss": 0.5164,
+      "step": 7320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2108949223077898,
+      "learning_rate": 8.2934951372672e-06,
+      "loss": 0.5398,
+      "step": 7321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2023682629600465,
+      "learning_rate": 8.291019376882955e-06,
+      "loss": 0.5501,
+      "step": 7322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.190400246059741,
+      "learning_rate": 8.28854372439547e-06,
+      "loss": 0.5521,
+      "step": 7323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1737444398725967,
+      "learning_rate": 8.28606817996105e-06,
+      "loss": 0.5256,
+      "step": 7324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.237148163611668,
+      "learning_rate": 8.283592743735988e-06,
+      "loss": 0.5686,
+      "step": 7325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2128477227183603,
+      "learning_rate": 8.281117415876574e-06,
+      "loss": 0.5371,
+      "step": 7326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.254491461501668,
+      "learning_rate": 8.27864219653909e-06,
+      "loss": 0.533,
+      "step": 7327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.067887559440243,
+      "learning_rate": 8.276167085879798e-06,
+      "loss": 0.5263,
+      "step": 7328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2076692345693945,
+      "learning_rate": 8.273692084054974e-06,
+      "loss": 0.555,
+      "step": 7329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.268412593081061,
+      "learning_rate": 8.271217191220874e-06,
+      "loss": 0.6188,
+      "step": 7330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2103028141562966,
+      "learning_rate": 8.268742407533754e-06,
+      "loss": 0.58,
+      "step": 7331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1120631463934354,
+      "learning_rate": 8.26626773314986e-06,
+      "loss": 0.5115,
+      "step": 7332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1895547958041968,
+      "learning_rate": 8.263793168225425e-06,
+      "loss": 0.5811,
+      "step": 7333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2021514124023913,
+      "learning_rate": 8.261318712916685e-06,
+      "loss": 0.5762,
+      "step": 7334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1191333284522993,
+      "learning_rate": 8.258844367379866e-06,
+      "loss": 0.4982,
+      "step": 7335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.140360239733101,
+      "learning_rate": 8.256370131771185e-06,
+      "loss": 0.5017,
+      "step": 7336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.246165641217594,
+      "learning_rate": 8.253896006246858e-06,
+      "loss": 0.5016,
+      "step": 7337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.086722275641157,
+      "learning_rate": 8.25142199096308e-06,
+      "loss": 0.5661,
+      "step": 7338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1408019255240671,
+      "learning_rate": 8.248948086076052e-06,
+      "loss": 0.5342,
+      "step": 7339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1338437877922387,
+      "learning_rate": 8.24647429174197e-06,
+      "loss": 0.5049,
+      "step": 7340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1914391685065533,
+      "learning_rate": 8.244000608117009e-06,
+      "loss": 0.5201,
+      "step": 7341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.178209199603273,
+      "learning_rate": 8.241527035357354e-06,
+      "loss": 0.5986,
+      "step": 7342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.106167791296938,
+      "learning_rate": 8.239053573619164e-06,
+      "loss": 0.5394,
+      "step": 7343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0767330084479267,
+      "learning_rate": 8.23658022305861e-06,
+      "loss": 0.4816,
+      "step": 7344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1955369726101355,
+      "learning_rate": 8.234106983831846e-06,
+      "loss": 0.5693,
+      "step": 7345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1524646243046868,
+      "learning_rate": 8.231633856095015e-06,
+      "loss": 0.5273,
+      "step": 7346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2637895532261076,
+      "learning_rate": 8.22916084000427e-06,
+      "loss": 0.5234,
+      "step": 7347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1900154375629686,
+      "learning_rate": 8.22668793571573e-06,
+      "loss": 0.5517,
+      "step": 7348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1714373979501782,
+      "learning_rate": 8.22421514338553e-06,
+      "loss": 0.5775,
+      "step": 7349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0615243920512325,
+      "learning_rate": 8.221742463169794e-06,
+      "loss": 0.5222,
+      "step": 7350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2751697389898624,
+      "learning_rate": 8.219269895224627e-06,
+      "loss": 0.533,
+      "step": 7351
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.132269367371601,
+      "learning_rate": 8.216797439706142e-06,
+      "loss": 0.5235,
+      "step": 7352
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1282529953811817,
+      "learning_rate": 8.214325096770433e-06,
+      "loss": 0.4858,
+      "step": 7353
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1682322023204603,
+      "learning_rate": 8.211852866573591e-06,
+      "loss": 0.5311,
+      "step": 7354
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2195835197630178,
+      "learning_rate": 8.209380749271708e-06,
+      "loss": 0.5723,
+      "step": 7355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2121316251885463,
+      "learning_rate": 8.206908745020852e-06,
+      "loss": 0.5515,
+      "step": 7356
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3240574651302344,
+      "learning_rate": 8.204436853977105e-06,
+      "loss": 0.5828,
+      "step": 7357
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1867401804191295,
+      "learning_rate": 8.201965076296518e-06,
+      "loss": 0.6011,
+      "step": 7358
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1582329217123883,
+      "learning_rate": 8.199493412135152e-06,
+      "loss": 0.5356,
+      "step": 7359
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2143649908842091,
+      "learning_rate": 8.197021861649059e-06,
+      "loss": 0.5619,
+      "step": 7360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2721781676289097,
+      "learning_rate": 8.194550424994274e-06,
+      "loss": 0.586,
+      "step": 7361
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2765637737440574,
+      "learning_rate": 8.192079102326842e-06,
+      "loss": 0.5512,
+      "step": 7362
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2384502885164446,
+      "learning_rate": 8.189607893802779e-06,
+      "loss": 0.5519,
+      "step": 7363
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2799023309340236,
+      "learning_rate": 8.187136799578111e-06,
+      "loss": 0.6006,
+      "step": 7364
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2247462340665136,
+      "learning_rate": 8.18466581980885e-06,
+      "loss": 0.566,
+      "step": 7365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1231214946337613,
+      "learning_rate": 8.182194954651e-06,
+      "loss": 0.5144,
+      "step": 7366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.143971433206162,
+      "learning_rate": 8.179724204260567e-06,
+      "loss": 0.5299,
+      "step": 7367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1346617764409792,
+      "learning_rate": 8.17725356879353e-06,
+      "loss": 0.5151,
+      "step": 7368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.131069763325533,
+      "learning_rate": 8.174783048405882e-06,
+      "loss": 0.548,
+      "step": 7369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.220352781756787,
+      "learning_rate": 8.172312643253597e-06,
+      "loss": 0.5244,
+      "step": 7370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2574127699399982,
+      "learning_rate": 8.169842353492644e-06,
+      "loss": 0.5885,
+      "step": 7371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1841725360063617,
+      "learning_rate": 8.167372179278988e-06,
+      "loss": 0.5662,
+      "step": 7372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.137280826622926,
+      "learning_rate": 8.164902120768578e-06,
+      "loss": 0.5198,
+      "step": 7373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.263724324524932,
+      "learning_rate": 8.162432178117365e-06,
+      "loss": 0.6203,
+      "step": 7374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1550726695959108,
+      "learning_rate": 8.159962351481291e-06,
+      "loss": 0.5361,
+      "step": 7375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1247691871378398,
+      "learning_rate": 8.157492641016285e-06,
+      "loss": 0.5336,
+      "step": 7376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0661411165407342,
+      "learning_rate": 8.15502304687828e-06,
+      "loss": 0.4792,
+      "step": 7377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.106380558391724,
+      "learning_rate": 8.152553569223183e-06,
+      "loss": 0.5188,
+      "step": 7378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3108831619440742,
+      "learning_rate": 8.150084208206912e-06,
+      "loss": 0.5872,
+      "step": 7379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1911717418571253,
+      "learning_rate": 8.147614963985371e-06,
+      "loss": 0.5382,
+      "step": 7380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1009492722898906,
+      "learning_rate": 8.145145836714452e-06,
+      "loss": 0.4933,
+      "step": 7381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2646844453729982,
+      "learning_rate": 8.142676826550046e-06,
+      "loss": 0.5811,
+      "step": 7382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1039165477447856,
+      "learning_rate": 8.14020793364804e-06,
+      "loss": 0.535,
+      "step": 7383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1592916227317223,
+      "learning_rate": 8.137739158164297e-06,
+      "loss": 0.5634,
+      "step": 7384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2085740682704509,
+      "learning_rate": 8.135270500254692e-06,
+      "loss": 0.5594,
+      "step": 7385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1477347578874182,
+      "learning_rate": 8.132801960075077e-06,
+      "loss": 0.5134,
+      "step": 7386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.151745191391416,
+      "learning_rate": 8.130333537781309e-06,
+      "loss": 0.5551,
+      "step": 7387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1227198912567786,
+      "learning_rate": 8.127865233529235e-06,
+      "loss": 0.5144,
+      "step": 7388
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.147854196241342,
+      "learning_rate": 8.125397047474684e-06,
+      "loss": 0.5008,
+      "step": 7389
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2012506722757004,
+      "learning_rate": 8.122928979773491e-06,
+      "loss": 0.5657,
+      "step": 7390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1194352166673986,
+      "learning_rate": 8.120461030581474e-06,
+      "loss": 0.4959,
+      "step": 7391
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2512962772527103,
+      "learning_rate": 8.117993200054449e-06,
+      "loss": 0.563,
+      "step": 7392
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.107005261295845,
+      "learning_rate": 8.115525488348227e-06,
+      "loss": 0.4826,
+      "step": 7393
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0240724064775895,
+      "learning_rate": 8.1130578956186e-06,
+      "loss": 0.4592,
+      "step": 7394
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1889458938749113,
+      "learning_rate": 8.110590422021365e-06,
+      "loss": 0.5709,
+      "step": 7395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2442022335302863,
+      "learning_rate": 8.108123067712302e-06,
+      "loss": 0.5585,
+      "step": 7396
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2719706412681977,
+      "learning_rate": 8.105655832847193e-06,
+      "loss": 0.5882,
+      "step": 7397
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2631577355296888,
+      "learning_rate": 8.103188717581808e-06,
+      "loss": 0.5847,
+      "step": 7398
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2090703043189124,
+      "learning_rate": 8.100721722071899e-06,
+      "loss": 0.5252,
+      "step": 7399
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2898948063210043,
+      "learning_rate": 8.09825484647323e-06,
+      "loss": 0.5635,
+      "step": 7400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.212812043376425,
+      "learning_rate": 8.095788090941543e-06,
+      "loss": 0.4894,
+      "step": 7401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1037216129303111,
+      "learning_rate": 8.093321455632578e-06,
+      "loss": 0.5261,
+      "step": 7402
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.217128751264944,
+      "learning_rate": 8.09085494070207e-06,
+      "loss": 0.5909,
+      "step": 7403
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1675637860405172,
+      "learning_rate": 8.088388546305737e-06,
+      "loss": 0.5218,
+      "step": 7404
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1547107761292843,
+      "learning_rate": 8.085922272599297e-06,
+      "loss": 0.4862,
+      "step": 7405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1765822410193867,
+      "learning_rate": 8.08345611973846e-06,
+      "loss": 0.5368,
+      "step": 7406
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1904028497538757,
+      "learning_rate": 8.080990087878925e-06,
+      "loss": 0.5099,
+      "step": 7407
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1265825161366,
+      "learning_rate": 8.078524177176392e-06,
+      "loss": 0.5165,
+      "step": 7408
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.108522759409965,
+      "learning_rate": 8.076058387786536e-06,
+      "loss": 0.4836,
+      "step": 7409
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1494970299813894,
+      "learning_rate": 8.073592719865041e-06,
+      "loss": 0.4565,
+      "step": 7410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2419361843938974,
+      "learning_rate": 8.071127173567576e-06,
+      "loss": 0.573,
+      "step": 7411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1678633626386437,
+      "learning_rate": 8.068661749049805e-06,
+      "loss": 0.5478,
+      "step": 7412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2158832579714967,
+      "learning_rate": 8.066196446467385e-06,
+      "loss": 0.5682,
+      "step": 7413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.144539946955945,
+      "learning_rate": 8.063731265975955e-06,
+      "loss": 0.58,
+      "step": 7414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2938475604290853,
+      "learning_rate": 8.061266207731165e-06,
+      "loss": 0.5421,
+      "step": 7415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2323920825192576,
+      "learning_rate": 8.058801271888637e-06,
+      "loss": 0.5627,
+      "step": 7416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2274845812231219,
+      "learning_rate": 8.056336458604002e-06,
+      "loss": 0.5575,
+      "step": 7417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1157593568738184,
+      "learning_rate": 8.053871768032878e-06,
+      "loss": 0.5078,
+      "step": 7418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1072647548406465,
+      "learning_rate": 8.051407200330866e-06,
+      "loss": 0.5665,
+      "step": 7419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1389622522517782,
+      "learning_rate": 8.048942755653573e-06,
+      "loss": 0.4904,
+      "step": 7420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1711855067977095,
+      "learning_rate": 8.046478434156588e-06,
+      "loss": 0.5653,
+      "step": 7421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.287530902380584,
+      "learning_rate": 8.044014235995496e-06,
+      "loss": 0.5378,
+      "step": 7422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1900621680842285,
+      "learning_rate": 8.041550161325884e-06,
+      "loss": 0.5696,
+      "step": 7423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1822198630331715,
+      "learning_rate": 8.039086210303308e-06,
+      "loss": 0.5236,
+      "step": 7424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.107481939715341,
+      "learning_rate": 8.036622383083341e-06,
+      "loss": 0.5087,
+      "step": 7425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.188553693990787,
+      "learning_rate": 8.034158679821529e-06,
+      "loss": 0.5522,
+      "step": 7426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2020325593235863,
+      "learning_rate": 8.031695100673423e-06,
+      "loss": 0.5276,
+      "step": 7427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.094103456333685,
+      "learning_rate": 8.029231645794564e-06,
+      "loss": 0.5193,
+      "step": 7428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1391752775899786,
+      "learning_rate": 8.026768315340475e-06,
+      "loss": 0.5384,
+      "step": 7429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1254403524061307,
+      "learning_rate": 8.024305109466685e-06,
+      "loss": 0.5018,
+      "step": 7430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3492063149573306,
+      "learning_rate": 8.021842028328703e-06,
+      "loss": 0.621,
+      "step": 7431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3184535466221496,
+      "learning_rate": 8.01937907208204e-06,
+      "loss": 0.5571,
+      "step": 7432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1071027669372246,
+      "learning_rate": 8.016916240882202e-06,
+      "loss": 0.5073,
+      "step": 7433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.125009324776893,
+      "learning_rate": 8.014453534884666e-06,
+      "loss": 0.5359,
+      "step": 7434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1636805579817773,
+      "learning_rate": 8.011990954244926e-06,
+      "loss": 0.55,
+      "step": 7435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1802868867720506,
+      "learning_rate": 8.00952849911845e-06,
+      "loss": 0.4918,
+      "step": 7436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0579360664354809,
+      "learning_rate": 8.007066169660711e-06,
+      "loss": 0.5026,
+      "step": 7437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1517000113673141,
+      "learning_rate": 8.004603966027167e-06,
+      "loss": 0.5381,
+      "step": 7438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2140958391126289,
+      "learning_rate": 8.002141888373274e-06,
+      "loss": 0.6047,
+      "step": 7439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.206806297697629,
+      "learning_rate": 7.999679936854467e-06,
+      "loss": 0.6038,
+      "step": 7440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.123099682437124,
+      "learning_rate": 7.997218111626186e-06,
+      "loss": 0.5236,
+      "step": 7441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1292954668203181,
+      "learning_rate": 7.994756412843857e-06,
+      "loss": 0.5644,
+      "step": 7442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.243385073545735,
+      "learning_rate": 7.992294840662904e-06,
+      "loss": 0.521,
+      "step": 7443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1395898066772119,
+      "learning_rate": 7.989833395238736e-06,
+      "loss": 0.5561,
+      "step": 7444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1119903575801624,
+      "learning_rate": 7.987372076726758e-06,
+      "loss": 0.5033,
+      "step": 7445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1777031484737517,
+      "learning_rate": 7.984910885282361e-06,
+      "loss": 0.604,
+      "step": 7446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1155423943206333,
+      "learning_rate": 7.982449821060936e-06,
+      "loss": 0.4984,
+      "step": 7447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1368127108932402,
+      "learning_rate": 7.979988884217867e-06,
+      "loss": 0.5662,
+      "step": 7448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2103154707521777,
+      "learning_rate": 7.97752807490852e-06,
+      "loss": 0.5269,
+      "step": 7449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2905908944932365,
+      "learning_rate": 7.97506739328826e-06,
+      "loss": 0.5872,
+      "step": 7450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1726942123581245,
+      "learning_rate": 7.972606839512442e-06,
+      "loss": 0.5305,
+      "step": 7451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2484401507104799,
+      "learning_rate": 7.970146413736414e-06,
+      "loss": 0.5875,
+      "step": 7452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1762299552774345,
+      "learning_rate": 7.967686116115517e-06,
+      "loss": 0.5538,
+      "step": 7453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3523709740149084,
+      "learning_rate": 7.965225946805081e-06,
+      "loss": 0.5547,
+      "step": 7454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1544494528065696,
+      "learning_rate": 7.962765905960428e-06,
+      "loss": 0.5179,
+      "step": 7455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1057632108033297,
+      "learning_rate": 7.960305993736874e-06,
+      "loss": 0.5098,
+      "step": 7456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2342755724301664,
+      "learning_rate": 7.957846210289725e-06,
+      "loss": 0.6316,
+      "step": 7457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2917100017210297,
+      "learning_rate": 7.955386555774284e-06,
+      "loss": 0.5408,
+      "step": 7458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2149208755788323,
+      "learning_rate": 7.952927030345836e-06,
+      "loss": 0.5383,
+      "step": 7459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2376940083869323,
+      "learning_rate": 7.950467634159669e-06,
+      "loss": 0.6077,
+      "step": 7460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.34088503576488,
+      "learning_rate": 7.94800836737105e-06,
+      "loss": 0.5636,
+      "step": 7461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1595330410814864,
+      "learning_rate": 7.945549230135251e-06,
+      "loss": 0.5838,
+      "step": 7462
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1809877215465043,
+      "learning_rate": 7.94309022260753e-06,
+      "loss": 0.5433,
+      "step": 7463
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.119626989749777,
+      "learning_rate": 7.940631344943137e-06,
+      "loss": 0.4819,
+      "step": 7464
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2423304349607722,
+      "learning_rate": 7.93817259729731e-06,
+      "loss": 0.5268,
+      "step": 7465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1871286614706904,
+      "learning_rate": 7.935713979825285e-06,
+      "loss": 0.5322,
+      "step": 7466
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2254199884536756,
+      "learning_rate": 7.933255492682287e-06,
+      "loss": 0.5376,
+      "step": 7467
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0613239736128324,
+      "learning_rate": 7.93079713602353e-06,
+      "loss": 0.4779,
+      "step": 7468
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1185703389062083,
+      "learning_rate": 7.92833891000423e-06,
+      "loss": 0.5153,
+      "step": 7469
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1915358175896347,
+      "learning_rate": 7.925880814779583e-06,
+      "loss": 0.5657,
+      "step": 7470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0711968103154779,
+      "learning_rate": 7.923422850504777e-06,
+      "loss": 0.5555,
+      "step": 7471
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2811896612334384,
+      "learning_rate": 7.920965017335002e-06,
+      "loss": 0.5162,
+      "step": 7472
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2086162345584948,
+      "learning_rate": 7.918507315425432e-06,
+      "loss": 0.5665,
+      "step": 7473
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1265145279700568,
+      "learning_rate": 7.916049744931236e-06,
+      "loss": 0.5498,
+      "step": 7474
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1811039486164794,
+      "learning_rate": 7.91359230600757e-06,
+      "loss": 0.5302,
+      "step": 7475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0706571594859717,
+      "learning_rate": 7.911134998809585e-06,
+      "loss": 0.521,
+      "step": 7476
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0910320298504563,
+      "learning_rate": 7.908677823492424e-06,
+      "loss": 0.5727,
+      "step": 7477
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.159279694457198,
+      "learning_rate": 7.906220780211225e-06,
+      "loss": 0.5083,
+      "step": 7478
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2002011726552926,
+      "learning_rate": 7.90376386912111e-06,
+      "loss": 0.5253,
+      "step": 7479
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.268780953165951,
+      "learning_rate": 7.901307090377197e-06,
+      "loss": 0.5538,
+      "step": 7480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1283256859475594,
+      "learning_rate": 7.898850444134592e-06,
+      "loss": 0.5385,
+      "step": 7481
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1390025474982035,
+      "learning_rate": 7.8963939305484e-06,
+      "loss": 0.5325,
+      "step": 7482
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0603178841397687,
+      "learning_rate": 7.893937549773716e-06,
+      "loss": 0.5175,
+      "step": 7483
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0824843466774459,
+      "learning_rate": 7.891481301965618e-06,
+      "loss": 0.5412,
+      "step": 7484
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3265900661807983,
+      "learning_rate": 7.889025187279185e-06,
+      "loss": 0.5421,
+      "step": 7485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2074165099761256,
+      "learning_rate": 7.886569205869481e-06,
+      "loss": 0.5717,
+      "step": 7486
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0868792395484372,
+      "learning_rate": 7.884113357891566e-06,
+      "loss": 0.5636,
+      "step": 7487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2002539624254245,
+      "learning_rate": 7.881657643500495e-06,
+      "loss": 0.6008,
+      "step": 7488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2799451277564378,
+      "learning_rate": 7.879202062851303e-06,
+      "loss": 0.5787,
+      "step": 7489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1200855786121509,
+      "learning_rate": 7.876746616099031e-06,
+      "loss": 0.4872,
+      "step": 7490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0010039535124273,
+      "learning_rate": 7.874291303398696e-06,
+      "loss": 0.5124,
+      "step": 7491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.03330788324661,
+      "learning_rate": 7.871836124905316e-06,
+      "loss": 0.4874,
+      "step": 7492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1881740062706498,
+      "learning_rate": 7.869381080773906e-06,
+      "loss": 0.477,
+      "step": 7493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1375168558066568,
+      "learning_rate": 7.866926171159458e-06,
+      "loss": 0.5021,
+      "step": 7494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.023417232400103,
+      "learning_rate": 7.86447139621697e-06,
+      "loss": 0.5129,
+      "step": 7495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0473909387738125,
+      "learning_rate": 7.862016756101417e-06,
+      "loss": 0.5049,
+      "step": 7496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2152867138246457,
+      "learning_rate": 7.859562250967776e-06,
+      "loss": 0.5431,
+      "step": 7497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.178171611238214,
+      "learning_rate": 7.857107880971015e-06,
+      "loss": 0.529,
+      "step": 7498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1518408243591383,
+      "learning_rate": 7.854653646266089e-06,
+      "loss": 0.496,
+      "step": 7499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.138098596961056,
+      "learning_rate": 7.85219954700795e-06,
+      "loss": 0.5358,
+      "step": 7500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1159691736516886,
+      "learning_rate": 7.84974558335153e-06,
+      "loss": 0.4809,
+      "step": 7501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2103259603597454,
+      "learning_rate": 7.847291755451766e-06,
+      "loss": 0.5712,
+      "step": 7502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2463382493036335,
+      "learning_rate": 7.844838063463582e-06,
+      "loss": 0.5461,
+      "step": 7503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.197938216748105,
+      "learning_rate": 7.842384507541889e-06,
+      "loss": 0.549,
+      "step": 7504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.303995005773564,
+      "learning_rate": 7.839931087841595e-06,
+      "loss": 0.569,
+      "step": 7505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2191946734542334,
+      "learning_rate": 7.837477804517595e-06,
+      "loss": 0.5447,
+      "step": 7506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.103673440906869,
+      "learning_rate": 7.835024657724778e-06,
+      "loss": 0.5132,
+      "step": 7507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2174770848117127,
+      "learning_rate": 7.832571647618024e-06,
+      "loss": 0.5735,
+      "step": 7508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1819323471253134,
+      "learning_rate": 7.830118774352205e-06,
+      "loss": 0.513,
+      "step": 7509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2180005238849785,
+      "learning_rate": 7.827666038082185e-06,
+      "loss": 0.5557,
+      "step": 7510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1654775326795914,
+      "learning_rate": 7.825213438962812e-06,
+      "loss": 0.4963,
+      "step": 7511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0044061032353246,
+      "learning_rate": 7.822760977148936e-06,
+      "loss": 0.4914,
+      "step": 7512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.105309734432696,
+      "learning_rate": 7.820308652795393e-06,
+      "loss": 0.5098,
+      "step": 7513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0785365217672074,
+      "learning_rate": 7.81785646605701e-06,
+      "loss": 0.5751,
+      "step": 7514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2183900937497338,
+      "learning_rate": 7.81540441708861e-06,
+      "loss": 0.4992,
+      "step": 7515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2433800880463783,
+      "learning_rate": 7.812952506044996e-06,
+      "loss": 0.5612,
+      "step": 7516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1851945884704642,
+      "learning_rate": 7.810500733080974e-06,
+      "loss": 0.5334,
+      "step": 7517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.103708868070359,
+      "learning_rate": 7.80804909835134e-06,
+      "loss": 0.5132,
+      "step": 7518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2963949486997477,
+      "learning_rate": 7.805597602010873e-06,
+      "loss": 0.6022,
+      "step": 7519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2372414829456624,
+      "learning_rate": 7.803146244214355e-06,
+      "loss": 0.567,
+      "step": 7520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1047388369705629,
+      "learning_rate": 7.800695025116546e-06,
+      "loss": 0.4538,
+      "step": 7521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1197263240231736,
+      "learning_rate": 7.798243944872204e-06,
+      "loss": 0.5379,
+      "step": 7522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0979699063406962,
+      "learning_rate": 7.795793003636085e-06,
+      "loss": 0.5311,
+      "step": 7523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.167235025972854,
+      "learning_rate": 7.793342201562923e-06,
+      "loss": 0.5486,
+      "step": 7524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1880808965452814,
+      "learning_rate": 7.790891538807459e-06,
+      "loss": 0.5442,
+      "step": 7525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1404507125011358,
+      "learning_rate": 7.788441015524403e-06,
+      "loss": 0.5421,
+      "step": 7526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1725054761518772,
+      "learning_rate": 7.785990631868478e-06,
+      "loss": 0.5171,
+      "step": 7527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2991012400819353,
+      "learning_rate": 7.783540387994387e-06,
+      "loss": 0.5217,
+      "step": 7528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1395736970557997,
+      "learning_rate": 7.781090284056827e-06,
+      "loss": 0.54,
+      "step": 7529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1179872864493376,
+      "learning_rate": 7.778640320210487e-06,
+      "loss": 0.4949,
+      "step": 7530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1500296423035712,
+      "learning_rate": 7.776190496610043e-06,
+      "loss": 0.5148,
+      "step": 7531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2500664216514035,
+      "learning_rate": 7.773740813410165e-06,
+      "loss": 0.587,
+      "step": 7532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1721832887130992,
+      "learning_rate": 7.771291270765518e-06,
+      "loss": 0.4816,
+      "step": 7533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.199138278311463,
+      "learning_rate": 7.768841868830747e-06,
+      "loss": 0.524,
+      "step": 7534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2628818973011953,
+      "learning_rate": 7.766392607760508e-06,
+      "loss": 0.5892,
+      "step": 7535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1327899799081569,
+      "learning_rate": 7.76394348770942e-06,
+      "loss": 0.4763,
+      "step": 7536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.068678618160993,
+      "learning_rate": 7.76149450883212e-06,
+      "loss": 0.5075,
+      "step": 7537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.143850443030654,
+      "learning_rate": 7.759045671283219e-06,
+      "loss": 0.5496,
+      "step": 7538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1350588920203621,
+      "learning_rate": 7.756596975217327e-06,
+      "loss": 0.5346,
+      "step": 7539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1705322201973778,
+      "learning_rate": 7.754148420789047e-06,
+      "loss": 0.5059,
+      "step": 7540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2948119518177412,
+      "learning_rate": 7.751700008152959e-06,
+      "loss": 0.578,
+      "step": 7541
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0714561867561079,
+      "learning_rate": 7.749251737463648e-06,
+      "loss": 0.5123,
+      "step": 7542
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2774052503053566,
+      "learning_rate": 7.74680360887569e-06,
+      "loss": 0.5184,
+      "step": 7543
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1999482163541297,
+      "learning_rate": 7.744355622543643e-06,
+      "loss": 0.5446,
+      "step": 7544
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3017220517259611,
+      "learning_rate": 7.741907778622064e-06,
+      "loss": 0.578,
+      "step": 7545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0748018104255457,
+      "learning_rate": 7.739460077265502e-06,
+      "loss": 0.514,
+      "step": 7546
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2338023245555976,
+      "learning_rate": 7.737012518628482e-06,
+      "loss": 0.4957,
+      "step": 7547
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.218626602968466,
+      "learning_rate": 7.73456510286554e-06,
+      "loss": 0.5326,
+      "step": 7548
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3630973014133738,
+      "learning_rate": 7.732117830131189e-06,
+      "loss": 0.5607,
+      "step": 7549
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.120810496205146,
+      "learning_rate": 7.72967070057994e-06,
+      "loss": 0.5164,
+      "step": 7550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2480994558227565,
+      "learning_rate": 7.727223714366299e-06,
+      "loss": 0.553,
+      "step": 7551
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2262898524785526,
+      "learning_rate": 7.724776871644745e-06,
+      "loss": 0.5761,
+      "step": 7552
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1602232428837291,
+      "learning_rate": 7.722330172569767e-06,
+      "loss": 0.5497,
+      "step": 7553
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2029902890575266,
+      "learning_rate": 7.719883617295835e-06,
+      "loss": 0.5727,
+      "step": 7554
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2652610443875372,
+      "learning_rate": 7.717437205977414e-06,
+      "loss": 0.5182,
+      "step": 7555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1500395415598474,
+      "learning_rate": 7.714990938768964e-06,
+      "loss": 0.5374,
+      "step": 7556
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1985369803906565,
+      "learning_rate": 7.712544815824921e-06,
+      "loss": 0.5478,
+      "step": 7557
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1611700284622768,
+      "learning_rate": 7.710098837299726e-06,
+      "loss": 0.5886,
+      "step": 7558
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1845211009396022,
+      "learning_rate": 7.707653003347805e-06,
+      "loss": 0.5464,
+      "step": 7559
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.322771434871977,
+      "learning_rate": 7.705207314123577e-06,
+      "loss": 0.6593,
+      "step": 7560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.111559154986994,
+      "learning_rate": 7.702761769781454e-06,
+      "loss": 0.4898,
+      "step": 7561
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1570569392925638,
+      "learning_rate": 7.700316370475828e-06,
+      "loss": 0.4695,
+      "step": 7562
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2980350224071178,
+      "learning_rate": 7.697871116361099e-06,
+      "loss": 0.5757,
+      "step": 7563
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1888964122215295,
+      "learning_rate": 7.69542600759164e-06,
+      "loss": 0.5171,
+      "step": 7564
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1658745812969207,
+      "learning_rate": 7.692981044321826e-06,
+      "loss": 0.5855,
+      "step": 7565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2530777234983987,
+      "learning_rate": 7.690536226706028e-06,
+      "loss": 0.5078,
+      "step": 7566
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2676092534879186,
+      "learning_rate": 7.688091554898587e-06,
+      "loss": 0.5521,
+      "step": 7567
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2287932599234257,
+      "learning_rate": 7.685647029053857e-06,
+      "loss": 0.5336,
+      "step": 7568
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2328067909949798,
+      "learning_rate": 7.683202649326169e-06,
+      "loss": 0.5133,
+      "step": 7569
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1685160499649005,
+      "learning_rate": 7.68075841586985e-06,
+      "loss": 0.5128,
+      "step": 7570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1003401447128849,
+      "learning_rate": 7.678314328839223e-06,
+      "loss": 0.5109,
+      "step": 7571
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2123902005690965,
+      "learning_rate": 7.675870388388586e-06,
+      "loss": 0.5613,
+      "step": 7572
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1539914522631276,
+      "learning_rate": 7.673426594672243e-06,
+      "loss": 0.4736,
+      "step": 7573
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.170372622973876,
+      "learning_rate": 7.670982947844482e-06,
+      "loss": 0.522,
+      "step": 7574
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0466722320501818,
+      "learning_rate": 7.668539448059585e-06,
+      "loss": 0.4948,
+      "step": 7575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0414242971105694,
+      "learning_rate": 7.666096095471823e-06,
+      "loss": 0.4884,
+      "step": 7576
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.192964429166772,
+      "learning_rate": 7.663652890235452e-06,
+      "loss": 0.5673,
+      "step": 7577
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1525280304021817,
+      "learning_rate": 7.661209832504731e-06,
+      "loss": 0.5482,
+      "step": 7578
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1793854181408703,
+      "learning_rate": 7.658766922433898e-06,
+      "loss": 0.499,
+      "step": 7579
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2662259311779778,
+      "learning_rate": 7.656324160177187e-06,
+      "loss": 0.52,
+      "step": 7580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2434548205031937,
+      "learning_rate": 7.653881545888829e-06,
+      "loss": 0.5092,
+      "step": 7581
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2173877830410018,
+      "learning_rate": 7.65143907972303e-06,
+      "loss": 0.5567,
+      "step": 7582
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2011819978613083,
+      "learning_rate": 7.648996761834e-06,
+      "loss": 0.5432,
+      "step": 7583
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1359903945046632,
+      "learning_rate": 7.64655459237593e-06,
+      "loss": 0.5661,
+      "step": 7584
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2234511548835636,
+      "learning_rate": 7.644112571503014e-06,
+      "loss": 0.5574,
+      "step": 7585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1583801443031645,
+      "learning_rate": 7.641670699369429e-06,
+      "loss": 0.5709,
+      "step": 7586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1141114468943734,
+      "learning_rate": 7.639228976129337e-06,
+      "loss": 0.5244,
+      "step": 7587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1627786199918313,
+      "learning_rate": 7.636787401936899e-06,
+      "loss": 0.5644,
+      "step": 7588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.124749897600938,
+      "learning_rate": 7.634345976946265e-06,
+      "loss": 0.5353,
+      "step": 7589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0454095860942514,
+      "learning_rate": 7.631904701311574e-06,
+      "loss": 0.4875,
+      "step": 7590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0700293953146631,
+      "learning_rate": 7.62946357518696e-06,
+      "loss": 0.5329,
+      "step": 7591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1372642650099527,
+      "learning_rate": 7.627022598726539e-06,
+      "loss": 0.5555,
+      "step": 7592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0988647758590586,
+      "learning_rate": 7.624581772084425e-06,
+      "loss": 0.4828,
+      "step": 7593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1854360110688413,
+      "learning_rate": 7.622141095414717e-06,
+      "loss": 0.5528,
+      "step": 7594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0399068399115785,
+      "learning_rate": 7.619700568871511e-06,
+      "loss": 0.4943,
+      "step": 7595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2005507933656396,
+      "learning_rate": 7.617260192608892e-06,
+      "loss": 0.5032,
+      "step": 7596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2502070732260793,
+      "learning_rate": 7.614819966780926e-06,
+      "loss": 0.5761,
+      "step": 7597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2954981285502771,
+      "learning_rate": 7.6123798915416845e-06,
+      "loss": 0.5198,
+      "step": 7598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1211739753967689,
+      "learning_rate": 7.609939967045217e-06,
+      "loss": 0.5431,
+      "step": 7599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1132347766478696,
+      "learning_rate": 7.60750019344557e-06,
+      "loss": 0.5113,
+      "step": 7600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0685703550461578,
+      "learning_rate": 7.605060570896781e-06,
+      "loss": 0.5003,
+      "step": 7601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.253303549358541,
+      "learning_rate": 7.602621099552874e-06,
+      "loss": 0.5236,
+      "step": 7602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1137780937992867,
+      "learning_rate": 7.600181779567867e-06,
+      "loss": 0.5163,
+      "step": 7603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2922049856790265,
+      "learning_rate": 7.597742611095762e-06,
+      "loss": 0.5713,
+      "step": 7604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0113709194283491,
+      "learning_rate": 7.595303594290562e-06,
+      "loss": 0.509,
+      "step": 7605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2947273395954355,
+      "learning_rate": 7.592864729306253e-06,
+      "loss": 0.5899,
+      "step": 7606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1870841753871757,
+      "learning_rate": 7.5904260162968145e-06,
+      "loss": 0.5193,
+      "step": 7607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.249563808629943,
+      "learning_rate": 7.5879874554162124e-06,
+      "loss": 0.5446,
+      "step": 7608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1636638598823918,
+      "learning_rate": 7.585549046818405e-06,
+      "loss": 0.5522,
+      "step": 7609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1918571244386105,
+      "learning_rate": 7.583110790657342e-06,
+      "loss": 0.5686,
+      "step": 7610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2336789353609325,
+      "learning_rate": 7.580672687086967e-06,
+      "loss": 0.5429,
+      "step": 7611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1284992579946584,
+      "learning_rate": 7.578234736261208e-06,
+      "loss": 0.4427,
+      "step": 7612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1799169658282695,
+      "learning_rate": 7.575796938333986e-06,
+      "loss": 0.5426,
+      "step": 7613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1395142254746995,
+      "learning_rate": 7.573359293459206e-06,
+      "loss": 0.4955,
+      "step": 7614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0797373008007733,
+      "learning_rate": 7.5709218017907734e-06,
+      "loss": 0.4721,
+      "step": 7615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.157151824211684,
+      "learning_rate": 7.568484463482584e-06,
+      "loss": 0.5152,
+      "step": 7616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2051507318964612,
+      "learning_rate": 7.566047278688514e-06,
+      "loss": 0.5205,
+      "step": 7617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1926528162577588,
+      "learning_rate": 7.563610247562437e-06,
+      "loss": 0.5057,
+      "step": 7618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1477989964946236,
+      "learning_rate": 7.561173370258215e-06,
+      "loss": 0.5426,
+      "step": 7619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.160051231897204,
+      "learning_rate": 7.558736646929699e-06,
+      "loss": 0.5412,
+      "step": 7620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2793791067470277,
+      "learning_rate": 7.556300077730735e-06,
+      "loss": 0.5875,
+      "step": 7621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1494886816495486,
+      "learning_rate": 7.553863662815156e-06,
+      "loss": 0.5226,
+      "step": 7622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1578831866021364,
+      "learning_rate": 7.551427402336784e-06,
+      "loss": 0.4992,
+      "step": 7623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2472773942353286,
+      "learning_rate": 7.548991296449431e-06,
+      "loss": 0.547,
+      "step": 7624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1410487707247865,
+      "learning_rate": 7.546555345306904e-06,
+      "loss": 0.5297,
+      "step": 7625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2668888704418546,
+      "learning_rate": 7.544119549062998e-06,
+      "loss": 0.5275,
+      "step": 7626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3520743663159378,
+      "learning_rate": 7.541683907871494e-06,
+      "loss": 0.5752,
+      "step": 7627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1963845724787652,
+      "learning_rate": 7.539248421886169e-06,
+      "loss": 0.5619,
+      "step": 7628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2333444566912413,
+      "learning_rate": 7.536813091260786e-06,
+      "loss": 0.5273,
+      "step": 7629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1379288462365693,
+      "learning_rate": 7.5343779161491e-06,
+      "loss": 0.5302,
+      "step": 7630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.259515639940586,
+      "learning_rate": 7.531942896704859e-06,
+      "loss": 0.5402,
+      "step": 7631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1645478126600004,
+      "learning_rate": 7.529508033081796e-06,
+      "loss": 0.5634,
+      "step": 7632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.154329870691558,
+      "learning_rate": 7.5270733254336385e-06,
+      "loss": 0.5316,
+      "step": 7633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1067942298959335,
+      "learning_rate": 7.524638773914097e-06,
+      "loss": 0.5262,
+      "step": 7634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1012922455979763,
+      "learning_rate": 7.522204378676882e-06,
+      "loss": 0.4972,
+      "step": 7635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.127135052202406,
+      "learning_rate": 7.5197701398756895e-06,
+      "loss": 0.5698,
+      "step": 7636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0509799925318029,
+      "learning_rate": 7.517336057664204e-06,
+      "loss": 0.434,
+      "step": 7637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.13333891114564,
+      "learning_rate": 7.514902132196104e-06,
+      "loss": 0.4974,
+      "step": 7638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1210752479010047,
+      "learning_rate": 7.5124683636250495e-06,
+      "loss": 0.4977,
+      "step": 7639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.102641125424338,
+      "learning_rate": 7.5100347521047025e-06,
+      "loss": 0.4417,
+      "step": 7640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1227709091132043,
+      "learning_rate": 7.5076012977887095e-06,
+      "loss": 0.5072,
+      "step": 7641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1782862950009478,
+      "learning_rate": 7.505168000830708e-06,
+      "loss": 0.5614,
+      "step": 7642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1371494276985863,
+      "learning_rate": 7.502734861384319e-06,
+      "loss": 0.5376,
+      "step": 7643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2617740678028007,
+      "learning_rate": 7.500301879603161e-06,
+      "loss": 0.5574,
+      "step": 7644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2344276803279728,
+      "learning_rate": 7.497869055640843e-06,
+      "loss": 0.5284,
+      "step": 7645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1017987694555291,
+      "learning_rate": 7.495436389650962e-06,
+      "loss": 0.5071,
+      "step": 7646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2738314996164046,
+      "learning_rate": 7.493003881787106e-06,
+      "loss": 0.5595,
+      "step": 7647
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0699546382771965,
+      "learning_rate": 7.490571532202846e-06,
+      "loss": 0.4724,
+      "step": 7648
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2542344373893626,
+      "learning_rate": 7.488139341051751e-06,
+      "loss": 0.5412,
+      "step": 7649
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2142250472629428,
+      "learning_rate": 7.48570730848738e-06,
+      "loss": 0.5147,
+      "step": 7650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1245022838360708,
+      "learning_rate": 7.48327543466328e-06,
+      "loss": 0.5037,
+      "step": 7651
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0819864091636586,
+      "learning_rate": 7.480843719732989e-06,
+      "loss": 0.4783,
+      "step": 7652
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2437783377021394,
+      "learning_rate": 7.478412163850026e-06,
+      "loss": 0.511,
+      "step": 7653
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.213420758206791,
+      "learning_rate": 7.475980767167914e-06,
+      "loss": 0.5529,
+      "step": 7654
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.102253690267413,
+      "learning_rate": 7.473549529840157e-06,
+      "loss": 0.4977,
+      "step": 7655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2417763563430688,
+      "learning_rate": 7.471118452020256e-06,
+      "loss": 0.576,
+      "step": 7656
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2196627891123033,
+      "learning_rate": 7.4686875338616914e-06,
+      "loss": 0.5549,
+      "step": 7657
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1907378275920073,
+      "learning_rate": 7.466256775517948e-06,
+      "loss": 0.547,
+      "step": 7658
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1468677302241865,
+      "learning_rate": 7.463826177142483e-06,
+      "loss": 0.4931,
+      "step": 7659
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1262522721347055,
+      "learning_rate": 7.461395738888755e-06,
+      "loss": 0.4846,
+      "step": 7660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1475780151375445,
+      "learning_rate": 7.458965460910214e-06,
+      "loss": 0.5383,
+      "step": 7661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1554937209087068,
+      "learning_rate": 7.456535343360292e-06,
+      "loss": 0.5361,
+      "step": 7662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.028826727433479,
+      "learning_rate": 7.454105386392419e-06,
+      "loss": 0.4948,
+      "step": 7663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1928663468393321,
+      "learning_rate": 7.451675590160006e-06,
+      "loss": 0.5492,
+      "step": 7664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2017612407008635,
+      "learning_rate": 7.449245954816459e-06,
+      "loss": 0.5441,
+      "step": 7665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0750432959420864,
+      "learning_rate": 7.446816480515177e-06,
+      "loss": 0.5057,
+      "step": 7666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1238838593018492,
+      "learning_rate": 7.444387167409541e-06,
+      "loss": 0.4991,
+      "step": 7667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2414968237895796,
+      "learning_rate": 7.441958015652934e-06,
+      "loss": 0.5342,
+      "step": 7668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.188644660590536,
+      "learning_rate": 7.4395290253987105e-06,
+      "loss": 0.5184,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.170451762287841,
+      "learning_rate": 7.43710019680023e-06,
+      "loss": 0.5241,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2056543589290634,
+      "learning_rate": 7.434671530010839e-06,
+      "loss": 0.5559,
+      "step": 7671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2776988524557864,
+      "learning_rate": 7.4322430251838674e-06,
+      "loss": 0.4768,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2197474286239363,
+      "learning_rate": 7.4298146824726445e-06,
+      "loss": 0.5505,
+      "step": 7673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1540999656068396,
+      "learning_rate": 7.427386502030478e-06,
+      "loss": 0.4792,
+      "step": 7674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1754700004776857,
+      "learning_rate": 7.424958484010675e-06,
+      "loss": 0.5453,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2249731956196317,
+      "learning_rate": 7.42253062856653e-06,
+      "loss": 0.526,
+      "step": 7676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.231689816069384,
+      "learning_rate": 7.420102935851324e-06,
+      "loss": 0.5738,
+      "step": 7677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.189048861251189,
+      "learning_rate": 7.417675406018332e-06,
+      "loss": 0.5025,
+      "step": 7678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.143413792375082,
+      "learning_rate": 7.4152480392208105e-06,
+      "loss": 0.5292,
+      "step": 7679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.170508032483189,
+      "learning_rate": 7.412820835612016e-06,
+      "loss": 0.4964,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1980959329676009,
+      "learning_rate": 7.410393795345193e-06,
+      "loss": 0.5273,
+      "step": 7681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1215090630344111,
+      "learning_rate": 7.407966918573568e-06,
+      "loss": 0.5034,
+      "step": 7682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1825853338763397,
+      "learning_rate": 7.40554020545037e-06,
+      "loss": 0.5244,
+      "step": 7683
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1462150920348926,
+      "learning_rate": 7.403113656128799e-06,
+      "loss": 0.5359,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1330780770317208,
+      "learning_rate": 7.400687270762061e-06,
+      "loss": 0.5165,
+      "step": 7685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2432446569791638,
+      "learning_rate": 7.398261049503348e-06,
+      "loss": 0.5654,
+      "step": 7686
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.188720678054079,
+      "learning_rate": 7.395834992505837e-06,
+      "loss": 0.5894,
+      "step": 7687
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.203485707298391,
+      "learning_rate": 7.393409099922703e-06,
+      "loss": 0.5521,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3112134986211703,
+      "learning_rate": 7.390983371907095e-06,
+      "loss": 0.6046,
+      "step": 7689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3880458016025612,
+      "learning_rate": 7.388557808612169e-06,
+      "loss": 0.6076,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1220474174427975,
+      "learning_rate": 7.3861324101910635e-06,
+      "loss": 0.5573,
+      "step": 7691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2300879964132578,
+      "learning_rate": 7.383707176796903e-06,
+      "loss": 0.5501,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1069855545293743,
+      "learning_rate": 7.38128210858281e-06,
+      "loss": 0.5312,
+      "step": 7693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0154173418642058,
+      "learning_rate": 7.378857205701885e-06,
+      "loss": 0.4921,
+      "step": 7694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.230866092529174,
+      "learning_rate": 7.376432468307227e-06,
+      "loss": 0.5444,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1537083191952524,
+      "learning_rate": 7.3740078965519266e-06,
+      "loss": 0.5298,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1363854198101881,
+      "learning_rate": 7.371583490589053e-06,
+      "loss": 0.5462,
+      "step": 7697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.105611569510098,
+      "learning_rate": 7.369159250571678e-06,
+      "loss": 0.5163,
+      "step": 7698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.089266661850344,
+      "learning_rate": 7.3667351766528505e-06,
+      "loss": 0.5066,
+      "step": 7699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.076758140205487,
+      "learning_rate": 7.364311268985616e-06,
+      "loss": 0.5309,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1024403422297548,
+      "learning_rate": 7.36188752772301e-06,
+      "loss": 0.4984,
+      "step": 7701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2113851396661608,
+      "learning_rate": 7.359463953018056e-06,
+      "loss": 0.4815,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2617883810664767,
+      "learning_rate": 7.357040545023768e-06,
+      "loss": 0.5286,
+      "step": 7703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1882840629910512,
+      "learning_rate": 7.354617303893143e-06,
+      "loss": 0.5431,
+      "step": 7704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.17086377064702,
+      "learning_rate": 7.352194229779177e-06,
+      "loss": 0.5774,
+      "step": 7705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2108643048222572,
+      "learning_rate": 7.349771322834851e-06,
+      "loss": 0.5097,
+      "step": 7706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.096815445633583,
+      "learning_rate": 7.347348583213133e-06,
+      "loss": 0.5229,
+      "step": 7707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.146123046084517,
+      "learning_rate": 7.344926011066985e-06,
+      "loss": 0.4909,
+      "step": 7708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1098479216587311,
+      "learning_rate": 7.34250360654936e-06,
+      "loss": 0.5004,
+      "step": 7709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1521967907317086,
+      "learning_rate": 7.3400813698131905e-06,
+      "loss": 0.587,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1912626086085143,
+      "learning_rate": 7.337659301011408e-06,
+      "loss": 0.5536,
+      "step": 7711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2622791380364655,
+      "learning_rate": 7.335237400296927e-06,
+      "loss": 0.5728,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1782423350543982,
+      "learning_rate": 7.332815667822659e-06,
+      "loss": 0.6221,
+      "step": 7713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2107529530336583,
+      "learning_rate": 7.3303941037415024e-06,
+      "loss": 0.5269,
+      "step": 7714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2087518964592812,
+      "learning_rate": 7.327972708206334e-06,
+      "loss": 0.5397,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.237342936249032,
+      "learning_rate": 7.325551481370038e-06,
+      "loss": 0.5233,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.168068769886466,
+      "learning_rate": 7.323130423385473e-06,
+      "loss": 0.5806,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1706644545065703,
+      "learning_rate": 7.320709534405494e-06,
+      "loss": 0.4631,
+      "step": 7718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2245851845914941,
+      "learning_rate": 7.318288814582951e-06,
+      "loss": 0.506,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1291926987428518,
+      "learning_rate": 7.315868264070667e-06,
+      "loss": 0.5415,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1108798713598735,
+      "learning_rate": 7.313447883021469e-06,
+      "loss": 0.5213,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2668293532667123,
+      "learning_rate": 7.3110276715881645e-06,
+      "loss": 0.5411,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.316964277401386,
+      "learning_rate": 7.308607629923557e-06,
+      "loss": 0.5327,
+      "step": 7723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0794143706686312,
+      "learning_rate": 7.3061877581804395e-06,
+      "loss": 0.563,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2034152783066068,
+      "learning_rate": 7.303768056511583e-06,
+      "loss": 0.5531,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.272269479747504,
+      "learning_rate": 7.301348525069762e-06,
+      "loss": 0.5658,
+      "step": 7726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2227362463377844,
+      "learning_rate": 7.29892916400773e-06,
+      "loss": 0.49,
+      "step": 7727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.136811504971153,
+      "learning_rate": 7.296509973478235e-06,
+      "loss": 0.4989,
+      "step": 7728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2317801616979487,
+      "learning_rate": 7.2940909536340184e-06,
+      "loss": 0.5197,
+      "step": 7729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1958420521665303,
+      "learning_rate": 7.291672104627796e-06,
+      "loss": 0.531,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1883224349750479,
+      "learning_rate": 7.289253426612288e-06,
+      "loss": 0.5429,
+      "step": 7731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1940057131121007,
+      "learning_rate": 7.286834919740195e-06,
+      "loss": 0.5843,
+      "step": 7732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1652537657262145,
+      "learning_rate": 7.2844165841642135e-06,
+      "loss": 0.5263,
+      "step": 7733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1133999945238846,
+      "learning_rate": 7.281998420037026e-06,
+      "loss": 0.5187,
+      "step": 7734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.178772931811347,
+      "learning_rate": 7.279580427511297e-06,
+      "loss": 0.5398,
+      "step": 7735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.156470200474254,
+      "learning_rate": 7.277162606739694e-06,
+      "loss": 0.4845,
+      "step": 7736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2183695957529324,
+      "learning_rate": 7.274744957874862e-06,
+      "loss": 0.5565,
+      "step": 7737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1619184064631618,
+      "learning_rate": 7.27232748106944e-06,
+      "loss": 0.4968,
+      "step": 7738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2777065030320225,
+      "learning_rate": 7.269910176476062e-06,
+      "loss": 0.5697,
+      "step": 7739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0787274985325965,
+      "learning_rate": 7.267493044247338e-06,
+      "loss": 0.5213,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.240664478125496,
+      "learning_rate": 7.265076084535874e-06,
+      "loss": 0.5802,
+      "step": 7741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1855375738894567,
+      "learning_rate": 7.262659297494266e-06,
+      "loss": 0.497,
+      "step": 7742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1670912696562183,
+      "learning_rate": 7.2602426832751005e-06,
+      "loss": 0.4886,
+      "step": 7743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2025837857565378,
+      "learning_rate": 7.2578262420309534e-06,
+      "loss": 0.5525,
+      "step": 7744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.168066371550638,
+      "learning_rate": 7.25540997391438e-06,
+      "loss": 0.5167,
+      "step": 7745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1812902514346169,
+      "learning_rate": 7.252993879077936e-06,
+      "loss": 0.5359,
+      "step": 7746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1309621497380071,
+      "learning_rate": 7.250577957674159e-06,
+      "loss": 0.4906,
+      "step": 7747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1578495714685226,
+      "learning_rate": 7.24816220985558e-06,
+      "loss": 0.4922,
+      "step": 7748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1195265286033935,
+      "learning_rate": 7.245746635774724e-06,
+      "loss": 0.4856,
+      "step": 7749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2022404577392567,
+      "learning_rate": 7.2433312355840916e-06,
+      "loss": 0.5518,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1430677098621624,
+      "learning_rate": 7.2409160094361774e-06,
+      "loss": 0.5409,
+      "step": 7751
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2245437142079847,
+      "learning_rate": 7.2385009574834705e-06,
+      "loss": 0.5524,
+      "step": 7752
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.120899196849831,
+      "learning_rate": 7.236086079878446e-06,
+      "loss": 0.4832,
+      "step": 7753
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1531149142684882,
+      "learning_rate": 7.2336713767735725e-06,
+      "loss": 0.5234,
+      "step": 7754
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.161670095300928,
+      "learning_rate": 7.231256848321293e-06,
+      "loss": 0.4833,
+      "step": 7755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1041531771909487,
+      "learning_rate": 7.228842494674053e-06,
+      "loss": 0.4979,
+      "step": 7756
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1242680287644997,
+      "learning_rate": 7.226428315984283e-06,
+      "loss": 0.5051,
+      "step": 7757
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2279510413670842,
+      "learning_rate": 7.224014312404404e-06,
+      "loss": 0.5773,
+      "step": 7758
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1478084476273744,
+      "learning_rate": 7.221600484086826e-06,
+      "loss": 0.5261,
+      "step": 7759
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1540163993562722,
+      "learning_rate": 7.219186831183944e-06,
+      "loss": 0.5333,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2380839771328047,
+      "learning_rate": 7.216773353848141e-06,
+      "loss": 0.522,
+      "step": 7761
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1390615748439636,
+      "learning_rate": 7.214360052231797e-06,
+      "loss": 0.4815,
+      "step": 7762
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.103024101505859,
+      "learning_rate": 7.211946926487274e-06,
+      "loss": 0.5089,
+      "step": 7763
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0285385784133456,
+      "learning_rate": 7.209533976766928e-06,
+      "loss": 0.5471,
+      "step": 7764
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1227130957794027,
+      "learning_rate": 7.207121203223102e-06,
+      "loss": 0.5557,
+      "step": 7765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1430226561636287,
+      "learning_rate": 7.204708606008119e-06,
+      "loss": 0.5567,
+      "step": 7766
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1506392982502094,
+      "learning_rate": 7.202296185274302e-06,
+      "loss": 0.4934,
+      "step": 7767
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0627155085395406,
+      "learning_rate": 7.199883941173963e-06,
+      "loss": 0.5273,
+      "step": 7768
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1614702791474936,
+      "learning_rate": 7.197471873859399e-06,
+      "loss": 0.5168,
+      "step": 7769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2146122457335764,
+      "learning_rate": 7.195059983482897e-06,
+      "loss": 0.5171,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1708842349046111,
+      "learning_rate": 7.192648270196726e-06,
+      "loss": 0.5207,
+      "step": 7771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1807842083074005,
+      "learning_rate": 7.1902367341531536e-06,
+      "loss": 0.5832,
+      "step": 7772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.245341202274705,
+      "learning_rate": 7.187825375504431e-06,
+      "loss": 0.542,
+      "step": 7773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.217804811221884,
+      "learning_rate": 7.185414194402805e-06,
+      "loss": 0.569,
+      "step": 7774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2187131729430711,
+      "learning_rate": 7.183003191000505e-06,
+      "loss": 0.5518,
+      "step": 7775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2335917728304344,
+      "learning_rate": 7.180592365449742e-06,
+      "loss": 0.5684,
+      "step": 7776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1324912306254484,
+      "learning_rate": 7.17818171790273e-06,
+      "loss": 0.4846,
+      "step": 7777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1520112673321037,
+      "learning_rate": 7.175771248511664e-06,
+      "loss": 0.5322,
+      "step": 7778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2145664597784722,
+      "learning_rate": 7.173360957428732e-06,
+      "loss": 0.5466,
+      "step": 7779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2196088845604676,
+      "learning_rate": 7.170950844806109e-06,
+      "loss": 0.5069,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1349496610124286,
+      "learning_rate": 7.168540910795951e-06,
+      "loss": 0.5332,
+      "step": 7781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1940255311475914,
+      "learning_rate": 7.166131155550414e-06,
+      "loss": 0.5538,
+      "step": 7782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.190497279888636,
+      "learning_rate": 7.163721579221638e-06,
+      "loss": 0.5228,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0679824415220855,
+      "learning_rate": 7.161312181961754e-06,
+      "loss": 0.4807,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1247656366160539,
+      "learning_rate": 7.158902963922879e-06,
+      "loss": 0.5591,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1692783834076288,
+      "learning_rate": 7.156493925257114e-06,
+      "loss": 0.5372,
+      "step": 7786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2981125315920066,
+      "learning_rate": 7.154085066116558e-06,
+      "loss": 0.5805,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2085272151137993,
+      "learning_rate": 7.1516763866532955e-06,
+      "loss": 0.531,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0897060840576522,
+      "learning_rate": 7.1492678870194e-06,
+      "loss": 0.4923,
+      "step": 7789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1414289384763234,
+      "learning_rate": 7.146859567366933e-06,
+      "loss": 0.4814,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1963225441743155,
+      "learning_rate": 7.144451427847937e-06,
+      "loss": 0.5589,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1827548739909985,
+      "learning_rate": 7.142043468614455e-06,
+      "loss": 0.5394,
+      "step": 7792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1961650924286023,
+      "learning_rate": 7.139635689818514e-06,
+      "loss": 0.5794,
+      "step": 7793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1906948279382121,
+      "learning_rate": 7.137228091612132e-06,
+      "loss": 0.5365,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.090560677279326,
+      "learning_rate": 7.134820674147313e-06,
+      "loss": 0.5128,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.149568792022806,
+      "learning_rate": 7.132413437576043e-06,
+      "loss": 0.5389,
+      "step": 7796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.202005484763658,
+      "learning_rate": 7.130006382050305e-06,
+      "loss": 0.4747,
+      "step": 7797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1965656069866897,
+      "learning_rate": 7.127599507722074e-06,
+      "loss": 0.5338,
+      "step": 7798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3071009077536964,
+      "learning_rate": 7.1251928147433065e-06,
+      "loss": 0.5717,
+      "step": 7799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0713810842072236,
+      "learning_rate": 7.122786303265952e-06,
+      "loss": 0.4685,
+      "step": 7800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2521506405139966,
+      "learning_rate": 7.1203799734419375e-06,
+      "loss": 0.5251,
+      "step": 7801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.140593907180692,
+      "learning_rate": 7.117973825423193e-06,
+      "loss": 0.5511,
+      "step": 7802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2983479695031779,
+      "learning_rate": 7.115567859361629e-06,
+      "loss": 0.5779,
+      "step": 7803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2454885131048963,
+      "learning_rate": 7.1131620754091505e-06,
+      "loss": 0.5042,
+      "step": 7804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.321252928110738,
+      "learning_rate": 7.110756473717646e-06,
+      "loss": 0.5228,
+      "step": 7805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.13213143757669,
+      "learning_rate": 7.108351054438988e-06,
+      "loss": 0.5296,
+      "step": 7806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.249380101035593,
+      "learning_rate": 7.105945817725046e-06,
+      "loss": 0.4867,
+      "step": 7807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1128109658071885,
+      "learning_rate": 7.103540763727678e-06,
+      "loss": 0.5154,
+      "step": 7808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1646007343191,
+      "learning_rate": 7.101135892598725e-06,
+      "loss": 0.5527,
+      "step": 7809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2326544347269293,
+      "learning_rate": 7.098731204490023e-06,
+      "loss": 0.5327,
+      "step": 7810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1507014063673973,
+      "learning_rate": 7.096326699553384e-06,
+      "loss": 0.5387,
+      "step": 7811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.233105646630224,
+      "learning_rate": 7.093922377940622e-06,
+      "loss": 0.5323,
+      "step": 7812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1774480417859856,
+      "learning_rate": 7.091518239803533e-06,
+      "loss": 0.5185,
+      "step": 7813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2019897653712068,
+      "learning_rate": 7.089114285293906e-06,
+      "loss": 0.5744,
+      "step": 7814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1609124182228903,
+      "learning_rate": 7.086710514563513e-06,
+      "loss": 0.5477,
+      "step": 7815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1763191386871017,
+      "learning_rate": 7.084306927764113e-06,
+      "loss": 0.5639,
+      "step": 7816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2453669518181323,
+      "learning_rate": 7.08190352504746e-06,
+      "loss": 0.545,
+      "step": 7817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2543348013799753,
+      "learning_rate": 7.079500306565291e-06,
+      "loss": 0.5758,
+      "step": 7818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1122521177868907,
+      "learning_rate": 7.07709727246934e-06,
+      "loss": 0.5454,
+      "step": 7819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2221274405391953,
+      "learning_rate": 7.074694422911315e-06,
+      "loss": 0.5464,
+      "step": 7820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.19652939231342,
+      "learning_rate": 7.072291758042926e-06,
+      "loss": 0.5225,
+      "step": 7821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1074683769998526,
+      "learning_rate": 7.0698892780158625e-06,
+      "loss": 0.5044,
+      "step": 7822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1954839246906357,
+      "learning_rate": 7.067486982981804e-06,
+      "loss": 0.5531,
+      "step": 7823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0720141173187567,
+      "learning_rate": 7.065084873092425e-06,
+      "loss": 0.517,
+      "step": 7824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2354404824562912,
+      "learning_rate": 7.062682948499377e-06,
+      "loss": 0.5565,
+      "step": 7825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.254465234055037,
+      "learning_rate": 7.060281209354315e-06,
+      "loss": 0.5384,
+      "step": 7826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2194844502688962,
+      "learning_rate": 7.057879655808862e-06,
+      "loss": 0.5459,
+      "step": 7827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2146731437774914,
+      "learning_rate": 7.055478288014646e-06,
+      "loss": 0.5148,
+      "step": 7828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1701081753305558,
+      "learning_rate": 7.053077106123278e-06,
+      "loss": 0.5894,
+      "step": 7829
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2052994931275567,
+      "learning_rate": 7.0506761102863565e-06,
+      "loss": 0.5449,
+      "step": 7830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.196580999136246,
+      "learning_rate": 7.048275300655472e-06,
+      "loss": 0.521,
+      "step": 7831
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.13749348565217,
+      "learning_rate": 7.0458746773821915e-06,
+      "loss": 0.514,
+      "step": 7832
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1666929491806517,
+      "learning_rate": 7.043474240618086e-06,
+      "loss": 0.5396,
+      "step": 7833
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0910625683948953,
+      "learning_rate": 7.0410739905147065e-06,
+      "loss": 0.5458,
+      "step": 7834
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.204570545897053,
+      "learning_rate": 7.03867392722359e-06,
+      "loss": 0.5585,
+      "step": 7835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2072295979608385,
+      "learning_rate": 7.0362740508962705e-06,
+      "loss": 0.541,
+      "step": 7836
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0771549670798806,
+      "learning_rate": 7.033874361684256e-06,
+      "loss": 0.51,
+      "step": 7837
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1567993405588204,
+      "learning_rate": 7.031474859739058e-06,
+      "loss": 0.5428,
+      "step": 7838
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1498365638530126,
+      "learning_rate": 7.02907554521217e-06,
+      "loss": 0.5344,
+      "step": 7839
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0391374718715964,
+      "learning_rate": 7.026676418255068e-06,
+      "loss": 0.5073,
+      "step": 7840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3187526540141947,
+      "learning_rate": 7.024277479019227e-06,
+      "loss": 0.6021,
+      "step": 7841
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2710415351707958,
+      "learning_rate": 7.021878727656097e-06,
+      "loss": 0.5852,
+      "step": 7842
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2183561422262938,
+      "learning_rate": 7.0194801643171324e-06,
+      "loss": 0.5688,
+      "step": 7843
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0695383655431216,
+      "learning_rate": 7.017081789153759e-06,
+      "loss": 0.5091,
+      "step": 7844
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1687122644870616,
+      "learning_rate": 7.014683602317402e-06,
+      "loss": 0.5178,
+      "step": 7845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0709712653379195,
+      "learning_rate": 7.012285603959477e-06,
+      "loss": 0.4918,
+      "step": 7846
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1266700217142505,
+      "learning_rate": 7.009887794231371e-06,
+      "loss": 0.5317,
+      "step": 7847
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.191408151093514,
+      "learning_rate": 7.0074901732844795e-06,
+      "loss": 0.5349,
+      "step": 7848
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0818673570915007,
+      "learning_rate": 7.005092741270168e-06,
+      "loss": 0.4733,
+      "step": 7849
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1800136494251199,
+      "learning_rate": 7.0026954983398045e-06,
+      "loss": 0.5287,
+      "step": 7850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2905621677279835,
+      "learning_rate": 7.0002984446447416e-06,
+      "loss": 0.5673,
+      "step": 7851
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1364672405524396,
+      "learning_rate": 6.997901580336312e-06,
+      "loss": 0.5333,
+      "step": 7852
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1678508584222989,
+      "learning_rate": 6.995504905565844e-06,
+      "loss": 0.5066,
+      "step": 7853
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0731681541337377,
+      "learning_rate": 6.993108420484652e-06,
+      "loss": 0.4843,
+      "step": 7854
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1146248723545837,
+      "learning_rate": 6.990712125244039e-06,
+      "loss": 0.5617,
+      "step": 7855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2610995541872534,
+      "learning_rate": 6.9883160199952996e-06,
+      "loss": 0.5487,
+      "step": 7856
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1544294200517597,
+      "learning_rate": 6.985920104889703e-06,
+      "loss": 0.5068,
+      "step": 7857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4344190669718466,
+      "learning_rate": 6.983524380078523e-06,
+      "loss": 0.5083,
+      "step": 7858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.167468675197255,
+      "learning_rate": 6.9811288457130115e-06,
+      "loss": 0.5143,
+      "step": 7859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2997469985965413,
+      "learning_rate": 6.9787335019444105e-06,
+      "loss": 0.5593,
+      "step": 7860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2287374760095242,
+      "learning_rate": 6.976338348923955e-06,
+      "loss": 0.5573,
+      "step": 7861
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.194006911189079,
+      "learning_rate": 6.973943386802855e-06,
+      "loss": 0.4866,
+      "step": 7862
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1653919181246282,
+      "learning_rate": 6.971548615732324e-06,
+      "loss": 0.525,
+      "step": 7863
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1742388979261293,
+      "learning_rate": 6.9691540358635504e-06,
+      "loss": 0.5382,
+      "step": 7864
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.177622320054089,
+      "learning_rate": 6.96675964734772e-06,
+      "loss": 0.516,
+      "step": 7865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2093866076639326,
+      "learning_rate": 6.964365450336008e-06,
+      "loss": 0.5821,
+      "step": 7866
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.100729336404497,
+      "learning_rate": 6.961971444979563e-06,
+      "loss": 0.486,
+      "step": 7867
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0935850291497822,
+      "learning_rate": 6.959577631429535e-06,
+      "loss": 0.5151,
+      "step": 7868
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2244038629951342,
+      "learning_rate": 6.9571840098370566e-06,
+      "loss": 0.5779,
+      "step": 7869
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.031875478465976,
+      "learning_rate": 6.95479058035325e-06,
+      "loss": 0.4658,
+      "step": 7870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0577686092224674,
+      "learning_rate": 6.952397343129232e-06,
+      "loss": 0.4849,
+      "step": 7871
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1623036967423783,
+      "learning_rate": 6.950004298316086e-06,
+      "loss": 0.5704,
+      "step": 7872
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1364622055999398,
+      "learning_rate": 6.947611446064908e-06,
+      "loss": 0.5005,
+      "step": 7873
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1683675027894271,
+      "learning_rate": 6.945218786526764e-06,
+      "loss": 0.5791,
+      "step": 7874
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1427361794726965,
+      "learning_rate": 6.94282631985272e-06,
+      "loss": 0.5238,
+      "step": 7875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2391517058655634,
+      "learning_rate": 6.940434046193824e-06,
+      "loss": 0.5186,
+      "step": 7876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1726582262199163,
+      "learning_rate": 6.9380419657011146e-06,
+      "loss": 0.5252,
+      "step": 7877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2253517930283762,
+      "learning_rate": 6.9356500785256135e-06,
+      "loss": 0.55,
+      "step": 7878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1347181930740233,
+      "learning_rate": 6.933258384818329e-06,
+      "loss": 0.4905,
+      "step": 7879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2204177400778107,
+      "learning_rate": 6.930866884730265e-06,
+      "loss": 0.5451,
+      "step": 7880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1676987556021625,
+      "learning_rate": 6.92847557841241e-06,
+      "loss": 0.5668,
+      "step": 7881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1790499448575396,
+      "learning_rate": 6.926084466015744e-06,
+      "loss": 0.5273,
+      "step": 7882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1865000781361859,
+      "learning_rate": 6.92369354769122e-06,
+      "loss": 0.4774,
+      "step": 7883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.347984915198981,
+      "learning_rate": 6.921302823589793e-06,
+      "loss": 0.5311,
+      "step": 7884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0963504395531058,
+      "learning_rate": 6.918912293862403e-06,
+      "loss": 0.4951,
+      "step": 7885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1679240444647572,
+      "learning_rate": 6.916521958659977e-06,
+      "loss": 0.5576,
+      "step": 7886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.154004674806167,
+      "learning_rate": 6.914131818133432e-06,
+      "loss": 0.5403,
+      "step": 7887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1454715966941336,
+      "learning_rate": 6.911741872433664e-06,
+      "loss": 0.5142,
+      "step": 7888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1530003632795518,
+      "learning_rate": 6.909352121711563e-06,
+      "loss": 0.5231,
+      "step": 7889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1927150354099902,
+      "learning_rate": 6.906962566118009e-06,
+      "loss": 0.5091,
+      "step": 7890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2525211658326436,
+      "learning_rate": 6.904573205803864e-06,
+      "loss": 0.5635,
+      "step": 7891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2196444139388871,
+      "learning_rate": 6.902184040919989e-06,
+      "loss": 0.5389,
+      "step": 7892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1768412833466393,
+      "learning_rate": 6.899795071617214e-06,
+      "loss": 0.5252,
+      "step": 7893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0752564567898892,
+      "learning_rate": 6.897406298046372e-06,
+      "loss": 0.5161,
+      "step": 7894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2337644008425996,
+      "learning_rate": 6.895017720358275e-06,
+      "loss": 0.5328,
+      "step": 7895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1659156334717273,
+      "learning_rate": 6.892629338703731e-06,
+      "loss": 0.4627,
+      "step": 7896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.114672998879971,
+      "learning_rate": 6.890241153233532e-06,
+      "loss": 0.51,
+      "step": 7897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0345112068413678,
+      "learning_rate": 6.88785316409845e-06,
+      "loss": 0.4291,
+      "step": 7898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2609283995270413,
+      "learning_rate": 6.885465371449254e-06,
+      "loss": 0.5425,
+      "step": 7899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1207460402538525,
+      "learning_rate": 6.883077775436697e-06,
+      "loss": 0.5067,
+      "step": 7900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0903690945125262,
+      "learning_rate": 6.880690376211522e-06,
+      "loss": 0.5209,
+      "step": 7901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1704934686920958,
+      "learning_rate": 6.878303173924461e-06,
+      "loss": 0.5338,
+      "step": 7902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2291753359127904,
+      "learning_rate": 6.875916168726225e-06,
+      "loss": 0.5307,
+      "step": 7903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0942555485341354,
+      "learning_rate": 6.8735293607675145e-06,
+      "loss": 0.465,
+      "step": 7904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1536306146288118,
+      "learning_rate": 6.8711427501990295e-06,
+      "loss": 0.5163,
+      "step": 7905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1635876909725476,
+      "learning_rate": 6.8687563371714425e-06,
+      "loss": 0.5227,
+      "step": 7906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.165310031175236,
+      "learning_rate": 6.866370121835428e-06,
+      "loss": 0.4596,
+      "step": 7907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1714660948382547,
+      "learning_rate": 6.863984104341634e-06,
+      "loss": 0.5202,
+      "step": 7908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1461271025024393,
+      "learning_rate": 6.861598284840701e-06,
+      "loss": 0.5232,
+      "step": 7909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0713591643967781,
+      "learning_rate": 6.859212663483261e-06,
+      "loss": 0.5011,
+      "step": 7910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2168841139379114,
+      "learning_rate": 6.8568272404199275e-06,
+      "loss": 0.5276,
+      "step": 7911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3146629222751425,
+      "learning_rate": 6.854442015801312e-06,
+      "loss": 0.5626,
+      "step": 7912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1516127514582708,
+      "learning_rate": 6.852056989778001e-06,
+      "loss": 0.5152,
+      "step": 7913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1675119176415645,
+      "learning_rate": 6.849672162500569e-06,
+      "loss": 0.5379,
+      "step": 7914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1914884945357567,
+      "learning_rate": 6.847287534119589e-06,
+      "loss": 0.574,
+      "step": 7915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0768226276045214,
+      "learning_rate": 6.844903104785611e-06,
+      "loss": 0.5135,
+      "step": 7916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1621777939089832,
+      "learning_rate": 6.842518874649182e-06,
+      "loss": 0.5081,
+      "step": 7917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0835137522678033,
+      "learning_rate": 6.840134843860826e-06,
+      "loss": 0.5088,
+      "step": 7918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2174607328940494,
+      "learning_rate": 6.837751012571059e-06,
+      "loss": 0.5342,
+      "step": 7919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1238253076628746,
+      "learning_rate": 6.835367380930384e-06,
+      "loss": 0.531,
+      "step": 7920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2086033629002453,
+      "learning_rate": 6.832983949089293e-06,
+      "loss": 0.5187,
+      "step": 7921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1637631741934549,
+      "learning_rate": 6.83060071719827e-06,
+      "loss": 0.5124,
+      "step": 7922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.31596208544439,
+      "learning_rate": 6.828217685407774e-06,
+      "loss": 0.5818,
+      "step": 7923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1935038627398995,
+      "learning_rate": 6.825834853868259e-06,
+      "loss": 0.5797,
+      "step": 7924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4366428473431188,
+      "learning_rate": 6.823452222730162e-06,
+      "loss": 0.5467,
+      "step": 7925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0911741169592177,
+      "learning_rate": 6.82106979214392e-06,
+      "loss": 0.4954,
+      "step": 7926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1608370443550886,
+      "learning_rate": 6.8186875622599434e-06,
+      "loss": 0.5174,
+      "step": 7927
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1283981605418782,
+      "learning_rate": 6.8163055332286344e-06,
+      "loss": 0.4814,
+      "step": 7928
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1954016558738603,
+      "learning_rate": 6.81392370520038e-06,
+      "loss": 0.4782,
+      "step": 7929
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2694381444975502,
+      "learning_rate": 6.81154207832556e-06,
+      "loss": 0.5992,
+      "step": 7930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1908383376058862,
+      "learning_rate": 6.809160652754539e-06,
+      "loss": 0.534,
+      "step": 7931
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.288604007447456,
+      "learning_rate": 6.806779428637671e-06,
+      "loss": 0.5975,
+      "step": 7932
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.227494584205629,
+      "learning_rate": 6.804398406125295e-06,
+      "loss": 0.5705,
+      "step": 7933
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2062536407573883,
+      "learning_rate": 6.802017585367728e-06,
+      "loss": 0.541,
+      "step": 7934
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.229015109314627,
+      "learning_rate": 6.799636966515293e-06,
+      "loss": 0.5743,
+      "step": 7935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2845677010798349,
+      "learning_rate": 6.797256549718287e-06,
+      "loss": 0.5599,
+      "step": 7936
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1578203311711954,
+      "learning_rate": 6.794876335127002e-06,
+      "loss": 0.5248,
+      "step": 7937
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2289615664730857,
+      "learning_rate": 6.79249632289171e-06,
+      "loss": 0.5245,
+      "step": 7938
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2706730329795046,
+      "learning_rate": 6.790116513162672e-06,
+      "loss": 0.5255,
+      "step": 7939
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2013442497314895,
+      "learning_rate": 6.787736906090139e-06,
+      "loss": 0.5322,
+      "step": 7940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2509952874312156,
+      "learning_rate": 6.7853575018243504e-06,
+      "loss": 0.5498,
+      "step": 7941
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1785243786672281,
+      "learning_rate": 6.7829783005155235e-06,
+      "loss": 0.5028,
+      "step": 7942
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2342227890232524,
+      "learning_rate": 6.780599302313882e-06,
+      "loss": 0.5354,
+      "step": 7943
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2008684969237855,
+      "learning_rate": 6.778220507369609e-06,
+      "loss": 0.5699,
+      "step": 7944
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0941171847212174,
+      "learning_rate": 6.7758419158328995e-06,
+      "loss": 0.5166,
+      "step": 7945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1501688978310545,
+      "learning_rate": 6.773463527853925e-06,
+      "loss": 0.4776,
+      "step": 7946
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1178643903667111,
+      "learning_rate": 6.7710853435828435e-06,
+      "loss": 0.5202,
+      "step": 7947
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1799499017781123,
+      "learning_rate": 6.768707363169809e-06,
+      "loss": 0.4878,
+      "step": 7948
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1902226306322088,
+      "learning_rate": 6.766329586764944e-06,
+      "loss": 0.5255,
+      "step": 7949
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.127517003779116,
+      "learning_rate": 6.7639520145183754e-06,
+      "loss": 0.4937,
+      "step": 7950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1889673503246165,
+      "learning_rate": 6.761574646580215e-06,
+      "loss": 0.5909,
+      "step": 7951
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.207067001455262,
+      "learning_rate": 6.759197483100553e-06,
+      "loss": 0.5352,
+      "step": 7952
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1140449982668799,
+      "learning_rate": 6.756820524229477e-06,
+      "loss": 0.49,
+      "step": 7953
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3042881520069007,
+      "learning_rate": 6.754443770117052e-06,
+      "loss": 0.607,
+      "step": 7954
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1481597460710176,
+      "learning_rate": 6.7520672209133355e-06,
+      "loss": 0.5398,
+      "step": 7955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1262324787601778,
+      "learning_rate": 6.749690876768374e-06,
+      "loss": 0.4918,
+      "step": 7956
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2529525695281212,
+      "learning_rate": 6.747314737832196e-06,
+      "loss": 0.5323,
+      "step": 7957
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1282482407478194,
+      "learning_rate": 6.744938804254823e-06,
+      "loss": 0.483,
+      "step": 7958
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.380644440278784,
+      "learning_rate": 6.742563076186255e-06,
+      "loss": 0.6295,
+      "step": 7959
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.180658559731154,
+      "learning_rate": 6.740187553776485e-06,
+      "loss": 0.5184,
+      "step": 7960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.167019359320177,
+      "learning_rate": 6.737812237175494e-06,
+      "loss": 0.5266,
+      "step": 7961
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2478725926550733,
+      "learning_rate": 6.735437126533246e-06,
+      "loss": 0.5071,
+      "step": 7962
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1053018612466925,
+      "learning_rate": 6.7330622219997e-06,
+      "loss": 0.4912,
+      "step": 7963
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1683931122005797,
+      "learning_rate": 6.730687523724787e-06,
+      "loss": 0.5673,
+      "step": 7964
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0695282227589669,
+      "learning_rate": 6.728313031858437e-06,
+      "loss": 0.4512,
+      "step": 7965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1487810827622318,
+      "learning_rate": 6.725938746550569e-06,
+      "loss": 0.4804,
+      "step": 7966
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2762308201109906,
+      "learning_rate": 6.723564667951076e-06,
+      "loss": 0.6093,
+      "step": 7967
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1882297883436632,
+      "learning_rate": 6.7211907962098535e-06,
+      "loss": 0.5255,
+      "step": 7968
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.170393248600238,
+      "learning_rate": 6.718817131476769e-06,
+      "loss": 0.5315,
+      "step": 7969
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1443256284604553,
+      "learning_rate": 6.7164436739016896e-06,
+      "loss": 0.5166,
+      "step": 7970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.211035004658763,
+      "learning_rate": 6.714070423634461e-06,
+      "loss": 0.5379,
+      "step": 7971
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2546069602722156,
+      "learning_rate": 6.7116973808249185e-06,
+      "loss": 0.5592,
+      "step": 7972
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.261041985287015,
+      "learning_rate": 6.709324545622891e-06,
+      "loss": 0.5618,
+      "step": 7973
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.253423199644568,
+      "learning_rate": 6.706951918178177e-06,
+      "loss": 0.5893,
+      "step": 7974
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.064863325402281,
+      "learning_rate": 6.70457949864058e-06,
+      "loss": 0.4969,
+      "step": 7975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0972214784633085,
+      "learning_rate": 6.702207287159882e-06,
+      "loss": 0.4627,
+      "step": 7976
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2453620221140367,
+      "learning_rate": 6.699835283885851e-06,
+      "loss": 0.5873,
+      "step": 7977
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2013890512033891,
+      "learning_rate": 6.697463488968249e-06,
+      "loss": 0.4977,
+      "step": 7978
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1240667074642812,
+      "learning_rate": 6.695091902556812e-06,
+      "loss": 0.5582,
+      "step": 7979
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1377490640802794,
+      "learning_rate": 6.692720524801273e-06,
+      "loss": 0.5416,
+      "step": 7980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.164344087887588,
+      "learning_rate": 6.690349355851353e-06,
+      "loss": 0.5033,
+      "step": 7981
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1565506518462345,
+      "learning_rate": 6.687978395856753e-06,
+      "loss": 0.5019,
+      "step": 7982
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1000411437749689,
+      "learning_rate": 6.6856076449671644e-06,
+      "loss": 0.468,
+      "step": 7983
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.229097504474967,
+      "learning_rate": 6.683237103332268e-06,
+      "loss": 0.5227,
+      "step": 7984
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1920046944909812,
+      "learning_rate": 6.6808667711017224e-06,
+      "loss": 0.5343,
+      "step": 7985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.185502882617292,
+      "learning_rate": 6.6784966484251845e-06,
+      "loss": 0.4656,
+      "step": 7986
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0356141427207093,
+      "learning_rate": 6.676126735452286e-06,
+      "loss": 0.4529,
+      "step": 7987
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0815745467490792,
+      "learning_rate": 6.673757032332657e-06,
+      "loss": 0.4856,
+      "step": 7988
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1164200287403079,
+      "learning_rate": 6.671387539215912e-06,
+      "loss": 0.5345,
+      "step": 7989
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2616897910370313,
+      "learning_rate": 6.66901825625164e-06,
+      "loss": 0.5522,
+      "step": 7990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2044423805158917,
+      "learning_rate": 6.666649183589435e-06,
+      "loss": 0.5221,
+      "step": 7991
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.242394915843983,
+      "learning_rate": 6.664280321378862e-06,
+      "loss": 0.513,
+      "step": 7992
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.290600315994042,
+      "learning_rate": 6.661911669769481e-06,
+      "loss": 0.5323,
+      "step": 7993
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.19674048800859,
+      "learning_rate": 6.659543228910844e-06,
+      "loss": 0.4963,
+      "step": 7994
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.183839778698033,
+      "learning_rate": 6.657174998952474e-06,
+      "loss": 0.5644,
+      "step": 7995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1439531969186578,
+      "learning_rate": 6.654806980043893e-06,
+      "loss": 0.5814,
+      "step": 7996
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2879789017287988,
+      "learning_rate": 6.6524391723346045e-06,
+      "loss": 0.5661,
+      "step": 7997
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1458643417786534,
+      "learning_rate": 6.650071575974104e-06,
+      "loss": 0.5082,
+      "step": 7998
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.183028082810128,
+      "learning_rate": 6.64770419111187e-06,
+      "loss": 0.5384,
+      "step": 7999
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3180446672989186,
+      "learning_rate": 6.645337017897364e-06,
+      "loss": 0.6019,
+      "step": 8000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1009780198933996,
+      "learning_rate": 6.64297005648004e-06,
+      "loss": 0.54,
+      "step": 8001
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1427698219097382,
+      "learning_rate": 6.640603307009337e-06,
+      "loss": 0.5164,
+      "step": 8002
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.146831609336497,
+      "learning_rate": 6.638236769634677e-06,
+      "loss": 0.4894,
+      "step": 8003
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1889526617221147,
+      "learning_rate": 6.635870444505479e-06,
+      "loss": 0.5086,
+      "step": 8004
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.435674088219694,
+      "learning_rate": 6.633504331771133e-06,
+      "loss": 0.5725,
+      "step": 8005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1108984896092007,
+      "learning_rate": 6.631138431581028e-06,
+      "loss": 0.5025,
+      "step": 8006
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1144714418143478,
+      "learning_rate": 6.628772744084534e-06,
+      "loss": 0.5254,
+      "step": 8007
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2043269705190347,
+      "learning_rate": 6.62640726943101e-06,
+      "loss": 0.5023,
+      "step": 8008
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1540997073768196,
+      "learning_rate": 6.624042007769804e-06,
+      "loss": 0.5433,
+      "step": 8009
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1627167468896367,
+      "learning_rate": 6.621676959250239e-06,
+      "loss": 0.5499,
+      "step": 8010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2574048062766159,
+      "learning_rate": 6.619312124021641e-06,
+      "loss": 0.5617,
+      "step": 8011
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.07319820129949,
+      "learning_rate": 6.6169475022333075e-06,
+      "loss": 0.5464,
+      "step": 8012
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.224178694779988,
+      "learning_rate": 6.614583094034533e-06,
+      "loss": 0.5208,
+      "step": 8013
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.146619226690463,
+      "learning_rate": 6.612218899574598e-06,
+      "loss": 0.539,
+      "step": 8014
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0959674156071852,
+      "learning_rate": 6.6098549190027584e-06,
+      "loss": 0.4975,
+      "step": 8015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1302351999678646,
+      "learning_rate": 6.60749115246827e-06,
+      "loss": 0.5423,
+      "step": 8016
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1742665619294352,
+      "learning_rate": 6.605127600120368e-06,
+      "loss": 0.4982,
+      "step": 8017
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.115350827418697,
+      "learning_rate": 6.602764262108274e-06,
+      "loss": 0.5443,
+      "step": 8018
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1259885788976922,
+      "learning_rate": 6.600401138581205e-06,
+      "loss": 0.532,
+      "step": 8019
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1503542520313583,
+      "learning_rate": 6.598038229688347e-06,
+      "loss": 0.5526,
+      "step": 8020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2370709301558145,
+      "learning_rate": 6.5956755355788895e-06,
+      "loss": 0.6149,
+      "step": 8021
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0876145620403093,
+      "learning_rate": 6.5933130564019976e-06,
+      "loss": 0.5004,
+      "step": 8022
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1713440264371344,
+      "learning_rate": 6.590950792306829e-06,
+      "loss": 0.5335,
+      "step": 8023
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.180905452312604,
+      "learning_rate": 6.5885887434425275e-06,
+      "loss": 0.5437,
+      "step": 8024
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.204417686082076,
+      "learning_rate": 6.586226909958218e-06,
+      "loss": 0.5385,
+      "step": 8025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0626486505977926,
+      "learning_rate": 6.583865292003015e-06,
+      "loss": 0.5315,
+      "step": 8026
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0937478746666014,
+      "learning_rate": 6.581503889726022e-06,
+      "loss": 0.5229,
+      "step": 8027
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2301821904819097,
+      "learning_rate": 6.579142703276325e-06,
+      "loss": 0.5037,
+      "step": 8028
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.099661844468642,
+      "learning_rate": 6.576781732803001e-06,
+      "loss": 0.5408,
+      "step": 8029
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.131593245389108,
+      "learning_rate": 6.574420978455105e-06,
+      "loss": 0.538,
+      "step": 8030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1688908537987979,
+      "learning_rate": 6.572060440381688e-06,
+      "loss": 0.5571,
+      "step": 8031
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.103406080362045,
+      "learning_rate": 6.569700118731779e-06,
+      "loss": 0.4678,
+      "step": 8032
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1022211364856964,
+      "learning_rate": 6.5673400136544e-06,
+      "loss": 0.5141,
+      "step": 8033
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1820127799638995,
+      "learning_rate": 6.564980125298559e-06,
+      "loss": 0.5604,
+      "step": 8034
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.158854000934489,
+      "learning_rate": 6.562620453813242e-06,
+      "loss": 0.5273,
+      "step": 8035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2873508996645737,
+      "learning_rate": 6.560260999347432e-06,
+      "loss": 0.4882,
+      "step": 8036
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1934890801449982,
+      "learning_rate": 6.557901762050091e-06,
+      "loss": 0.5406,
+      "step": 8037
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2175066547877735,
+      "learning_rate": 6.555542742070169e-06,
+      "loss": 0.5582,
+      "step": 8038
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2787965684393074,
+      "learning_rate": 6.553183939556608e-06,
+      "loss": 0.5725,
+      "step": 8039
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1375518053435603,
+      "learning_rate": 6.550825354658328e-06,
+      "loss": 0.5195,
+      "step": 8040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1587409946457963,
+      "learning_rate": 6.548466987524238e-06,
+      "loss": 0.5572,
+      "step": 8041
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3383814002051737,
+      "learning_rate": 6.546108838303233e-06,
+      "loss": 0.517,
+      "step": 8042
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1342899553350756,
+      "learning_rate": 6.543750907144196e-06,
+      "loss": 0.5599,
+      "step": 8043
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2611076835738857,
+      "learning_rate": 6.5413931941959994e-06,
+      "loss": 0.564,
+      "step": 8044
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1934999174030916,
+      "learning_rate": 6.539035699607494e-06,
+      "loss": 0.538,
+      "step": 8045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1913859881569846,
+      "learning_rate": 6.53667842352752e-06,
+      "loss": 0.5237,
+      "step": 8046
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.142456987268308,
+      "learning_rate": 6.534321366104905e-06,
+      "loss": 0.5534,
+      "step": 8047
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2650165743740918,
+      "learning_rate": 6.531964527488463e-06,
+      "loss": 0.5439,
+      "step": 8048
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.177078241432902,
+      "learning_rate": 6.529607907826994e-06,
+      "loss": 0.5396,
+      "step": 8049
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.293195952307599,
+      "learning_rate": 6.527251507269283e-06,
+      "loss": 0.598,
+      "step": 8050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2534863966366128,
+      "learning_rate": 6.524895325964102e-06,
+      "loss": 0.5722,
+      "step": 8051
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1097887906184716,
+      "learning_rate": 6.522539364060205e-06,
+      "loss": 0.5223,
+      "step": 8052
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1559349352500576,
+      "learning_rate": 6.5201836217063395e-06,
+      "loss": 0.5119,
+      "step": 8053
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.160296653189349,
+      "learning_rate": 6.517828099051238e-06,
+      "loss": 0.5649,
+      "step": 8054
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0796615045401898,
+      "learning_rate": 6.515472796243615e-06,
+      "loss": 0.4854,
+      "step": 8055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.180626754234388,
+      "learning_rate": 6.513117713432171e-06,
+      "loss": 0.5219,
+      "step": 8056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2019430522725991,
+      "learning_rate": 6.510762850765594e-06,
+      "loss": 0.4822,
+      "step": 8057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3580501282615738,
+      "learning_rate": 6.50840820839256e-06,
+      "loss": 0.5589,
+      "step": 8058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0983115307134994,
+      "learning_rate": 6.506053786461732e-06,
+      "loss": 0.5241,
+      "step": 8059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1142515003450155,
+      "learning_rate": 6.5036995851217565e-06,
+      "loss": 0.5246,
+      "step": 8060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.135803588362028,
+      "learning_rate": 6.501345604521263e-06,
+      "loss": 0.5331,
+      "step": 8061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1589763048364812,
+      "learning_rate": 6.49899184480887e-06,
+      "loss": 0.5037,
+      "step": 8062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1659016769169397,
+      "learning_rate": 6.496638306133186e-06,
+      "loss": 0.5319,
+      "step": 8063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1396461885588618,
+      "learning_rate": 6.494284988642803e-06,
+      "loss": 0.489,
+      "step": 8064
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.190222981182132,
+      "learning_rate": 6.491931892486294e-06,
+      "loss": 0.5518,
+      "step": 8065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1799287865049926,
+      "learning_rate": 6.489579017812224e-06,
+      "loss": 0.5291,
+      "step": 8066
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2032853552234715,
+      "learning_rate": 6.487226364769141e-06,
+      "loss": 0.5397,
+      "step": 8067
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1988851971805627,
+      "learning_rate": 6.484873933505581e-06,
+      "loss": 0.5361,
+      "step": 8068
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1916125009582101,
+      "learning_rate": 6.482521724170068e-06,
+      "loss": 0.5381,
+      "step": 8069
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1859034295898843,
+      "learning_rate": 6.480169736911104e-06,
+      "loss": 0.5412,
+      "step": 8070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1138086508604188,
+      "learning_rate": 6.477817971877186e-06,
+      "loss": 0.5013,
+      "step": 8071
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2226249032679162,
+      "learning_rate": 6.47546642921679e-06,
+      "loss": 0.4959,
+      "step": 8072
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1791949222715465,
+      "learning_rate": 6.4731151090783814e-06,
+      "loss": 0.498,
+      "step": 8073
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1192722209899926,
+      "learning_rate": 6.470764011610415e-06,
+      "loss": 0.544,
+      "step": 8074
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.255417480561332,
+      "learning_rate": 6.468413136961325e-06,
+      "loss": 0.5461,
+      "step": 8075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2249791805288721,
+      "learning_rate": 6.4660624852795336e-06,
+      "loss": 0.5201,
+      "step": 8076
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2620239361126833,
+      "learning_rate": 6.463712056713449e-06,
+      "loss": 0.568,
+      "step": 8077
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.201600682082281,
+      "learning_rate": 6.461361851411466e-06,
+      "loss": 0.5274,
+      "step": 8078
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.294499162864488,
+      "learning_rate": 6.45901186952197e-06,
+      "loss": 0.5912,
+      "step": 8079
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1582935934774032,
+      "learning_rate": 6.456662111193322e-06,
+      "loss": 0.5443,
+      "step": 8080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2203371521710187,
+      "learning_rate": 6.454312576573878e-06,
+      "loss": 0.567,
+      "step": 8081
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.172150388784612,
+      "learning_rate": 6.451963265811971e-06,
+      "loss": 0.4928,
+      "step": 8082
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1801006780929693,
+      "learning_rate": 6.449614179055929e-06,
+      "loss": 0.563,
+      "step": 8083
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1425055065064067,
+      "learning_rate": 6.447265316454063e-06,
+      "loss": 0.5077,
+      "step": 8084
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1370770915358563,
+      "learning_rate": 6.444916678154667e-06,
+      "loss": 0.4995,
+      "step": 8085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.104175687549137,
+      "learning_rate": 6.442568264306024e-06,
+      "loss": 0.5123,
+      "step": 8086
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2270560273962663,
+      "learning_rate": 6.440220075056398e-06,
+      "loss": 0.5398,
+      "step": 8087
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.241972082880874,
+      "learning_rate": 6.437872110554044e-06,
+      "loss": 0.5909,
+      "step": 8088
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1515595952363562,
+      "learning_rate": 6.4355243709472045e-06,
+      "loss": 0.5595,
+      "step": 8089
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0947170342227233,
+      "learning_rate": 6.433176856384103e-06,
+      "loss": 0.5011,
+      "step": 8090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1940678118517964,
+      "learning_rate": 6.430829567012946e-06,
+      "loss": 0.5401,
+      "step": 8091
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2344310120054822,
+      "learning_rate": 6.428482502981933e-06,
+      "loss": 0.5629,
+      "step": 8092
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1614183438976375,
+      "learning_rate": 6.426135664439246e-06,
+      "loss": 0.5458,
+      "step": 8093
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0956054752784061,
+      "learning_rate": 6.423789051533056e-06,
+      "loss": 0.5425,
+      "step": 8094
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.221843266379924,
+      "learning_rate": 6.4214426644115104e-06,
+      "loss": 0.5249,
+      "step": 8095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1928426120026385,
+      "learning_rate": 6.419096503222757e-06,
+      "loss": 0.5446,
+      "step": 8096
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.228508492890292,
+      "learning_rate": 6.416750568114911e-06,
+      "loss": 0.5188,
+      "step": 8097
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.192796490116827,
+      "learning_rate": 6.414404859236091e-06,
+      "loss": 0.5588,
+      "step": 8098
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2827298526880522,
+      "learning_rate": 6.412059376734392e-06,
+      "loss": 0.5426,
+      "step": 8099
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1082316672273647,
+      "learning_rate": 6.409714120757895e-06,
+      "loss": 0.5164,
+      "step": 8100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2100529547208914,
+      "learning_rate": 6.407369091454672e-06,
+      "loss": 0.5197,
+      "step": 8101
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2260628915838525,
+      "learning_rate": 6.40502428897277e-06,
+      "loss": 0.5549,
+      "step": 8102
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1489381801651146,
+      "learning_rate": 6.402679713460234e-06,
+      "loss": 0.5005,
+      "step": 8103
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.174123412707984,
+      "learning_rate": 6.400335365065087e-06,
+      "loss": 0.5305,
+      "step": 8104
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2181456118738843,
+      "learning_rate": 6.397991243935339e-06,
+      "loss": 0.5451,
+      "step": 8105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1271746069052053,
+      "learning_rate": 6.395647350218992e-06,
+      "loss": 0.4685,
+      "step": 8106
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1139567149740126,
+      "learning_rate": 6.393303684064019e-06,
+      "loss": 0.5267,
+      "step": 8107
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0985159985640889,
+      "learning_rate": 6.390960245618394e-06,
+      "loss": 0.4817,
+      "step": 8108
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1378639980229739,
+      "learning_rate": 6.388617035030069e-06,
+      "loss": 0.577,
+      "step": 8109
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2483777964117484,
+      "learning_rate": 6.386274052446982e-06,
+      "loss": 0.5169,
+      "step": 8110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1303506869951976,
+      "learning_rate": 6.383931298017063e-06,
+      "loss": 0.4857,
+      "step": 8111
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1579573114726083,
+      "learning_rate": 6.381588771888213e-06,
+      "loss": 0.512,
+      "step": 8112
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0717487583389012,
+      "learning_rate": 6.379246474208332e-06,
+      "loss": 0.5091,
+      "step": 8113
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1624755795282637,
+      "learning_rate": 6.376904405125304e-06,
+      "loss": 0.5323,
+      "step": 8114
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.143481140638387,
+      "learning_rate": 6.374562564786993e-06,
+      "loss": 0.5195,
+      "step": 8115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1233260628684765,
+      "learning_rate": 6.372220953341254e-06,
+      "loss": 0.5013,
+      "step": 8116
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.080642911234163,
+      "learning_rate": 6.36987957093592e-06,
+      "loss": 0.5109,
+      "step": 8117
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1621251209378372,
+      "learning_rate": 6.367538417718817e-06,
+      "loss": 0.4952,
+      "step": 8118
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2936428652278116,
+      "learning_rate": 6.365197493837757e-06,
+      "loss": 0.5466,
+      "step": 8119
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0631700534681894,
+      "learning_rate": 6.362856799440531e-06,
+      "loss": 0.517,
+      "step": 8120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1429910549404527,
+      "learning_rate": 6.360516334674924e-06,
+      "loss": 0.518,
+      "step": 8121
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2255141522179076,
+      "learning_rate": 6.358176099688693e-06,
+      "loss": 0.5635,
+      "step": 8122
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1898499877525728,
+      "learning_rate": 6.355836094629596e-06,
+      "loss": 0.5652,
+      "step": 8123
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2327496898204358,
+      "learning_rate": 6.353496319645369e-06,
+      "loss": 0.5372,
+      "step": 8124
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2067161568526497,
+      "learning_rate": 6.351156774883731e-06,
+      "loss": 0.554,
+      "step": 8125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1636667282846072,
+      "learning_rate": 6.348817460492396e-06,
+      "loss": 0.4822,
+      "step": 8126
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1580437330213513,
+      "learning_rate": 6.346478376619046e-06,
+      "loss": 0.4633,
+      "step": 8127
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1926301766764904,
+      "learning_rate": 6.344139523411368e-06,
+      "loss": 0.514,
+      "step": 8128
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1422670120137954,
+      "learning_rate": 6.341800901017024e-06,
+      "loss": 0.5179,
+      "step": 8129
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.086437467019316,
+      "learning_rate": 6.339462509583663e-06,
+      "loss": 0.4955,
+      "step": 8130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1578803038724266,
+      "learning_rate": 6.337124349258923e-06,
+      "loss": 0.5505,
+      "step": 8131
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2396365189028185,
+      "learning_rate": 6.334786420190415e-06,
+      "loss": 0.5078,
+      "step": 8132
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1917749553553898,
+      "learning_rate": 6.3324487225257526e-06,
+      "loss": 0.5211,
+      "step": 8133
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1814016053131546,
+      "learning_rate": 6.330111256412525e-06,
+      "loss": 0.5064,
+      "step": 8134
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.173899516697287,
+      "learning_rate": 6.3277740219983066e-06,
+      "loss": 0.5611,
+      "step": 8135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0959501209081624,
+      "learning_rate": 6.325437019430665e-06,
+      "loss": 0.4615,
+      "step": 8136
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1833449501935949,
+      "learning_rate": 6.323100248857137e-06,
+      "loss": 0.5287,
+      "step": 8137
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1935617927660638,
+      "learning_rate": 6.320763710425262e-06,
+      "loss": 0.5522,
+      "step": 8138
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.246307593863828,
+      "learning_rate": 6.318427404282557e-06,
+      "loss": 0.5709,
+      "step": 8139
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1493925936888927,
+      "learning_rate": 6.316091330576523e-06,
+      "loss": 0.5546,
+      "step": 8140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2851754508859623,
+      "learning_rate": 6.313755489454654e-06,
+      "loss": 0.5924,
+      "step": 8141
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2259033764697531,
+      "learning_rate": 6.311419881064416e-06,
+      "loss": 0.5237,
+      "step": 8142
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1730304870578598,
+      "learning_rate": 6.309084505553269e-06,
+      "loss": 0.5347,
+      "step": 8143
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2374726552543034,
+      "learning_rate": 6.306749363068665e-06,
+      "loss": 0.5426,
+      "step": 8144
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1704444800648626,
+      "learning_rate": 6.304414453758024e-06,
+      "loss": 0.5033,
+      "step": 8145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1944052051535528,
+      "learning_rate": 6.3020797777687705e-06,
+      "loss": 0.5466,
+      "step": 8146
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2234306442438918,
+      "learning_rate": 6.299745335248295e-06,
+      "loss": 0.4611,
+      "step": 8147
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1135698446354996,
+      "learning_rate": 6.297411126343988e-06,
+      "loss": 0.4788,
+      "step": 8148
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2336560340508356,
+      "learning_rate": 6.295077151203221e-06,
+      "loss": 0.5432,
+      "step": 8149
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0181511786498405,
+      "learning_rate": 6.292743409973345e-06,
+      "loss": 0.4528,
+      "step": 8150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2724951317175053,
+      "learning_rate": 6.290409902801706e-06,
+      "loss": 0.5483,
+      "step": 8151
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1709874670496545,
+      "learning_rate": 6.288076629835633e-06,
+      "loss": 0.5352,
+      "step": 8152
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2000504880456666,
+      "learning_rate": 6.285743591222428e-06,
+      "loss": 0.5182,
+      "step": 8153
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1198374125281498,
+      "learning_rate": 6.283410787109396e-06,
+      "loss": 0.5429,
+      "step": 8154
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1693575458049734,
+      "learning_rate": 6.2810782176438145e-06,
+      "loss": 0.5388,
+      "step": 8155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2620790044022467,
+      "learning_rate": 6.278745882972952e-06,
+      "loss": 0.5676,
+      "step": 8156
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1385941388927983,
+      "learning_rate": 6.276413783244064e-06,
+      "loss": 0.5279,
+      "step": 8157
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2475136347543503,
+      "learning_rate": 6.274081918604382e-06,
+      "loss": 0.5334,
+      "step": 8158
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1627673938822398,
+      "learning_rate": 6.271750289201134e-06,
+      "loss": 0.5334,
+      "step": 8159
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1870628355342407,
+      "learning_rate": 6.269418895181523e-06,
+      "loss": 0.5608,
+      "step": 8160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.285418637282944,
+      "learning_rate": 6.267087736692744e-06,
+      "loss": 0.5837,
+      "step": 8161
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0852448703949569,
+      "learning_rate": 6.26475681388198e-06,
+      "loss": 0.5261,
+      "step": 8162
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1479191551019983,
+      "learning_rate": 6.262426126896386e-06,
+      "loss": 0.5194,
+      "step": 8163
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.198075286757198,
+      "learning_rate": 6.260095675883116e-06,
+      "loss": 0.5426,
+      "step": 8164
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5790903896674928,
+      "learning_rate": 6.257765460989298e-06,
+      "loss": 0.493,
+      "step": 8165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.193184747733611,
+      "learning_rate": 6.255435482362056e-06,
+      "loss": 0.5195,
+      "step": 8166
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.140465032778158,
+      "learning_rate": 6.253105740148493e-06,
+      "loss": 0.4906,
+      "step": 8167
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0894346392223409,
+      "learning_rate": 6.2507762344956925e-06,
+      "loss": 0.4609,
+      "step": 8168
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1849297263639302,
+      "learning_rate": 6.248446965550735e-06,
+      "loss": 0.513,
+      "step": 8169
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0231387452150094,
+      "learning_rate": 6.246117933460673e-06,
+      "loss": 0.4737,
+      "step": 8170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.376445660520037,
+      "learning_rate": 6.2437891383725535e-06,
+      "loss": 0.5844,
+      "step": 8171
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1341401311520276,
+      "learning_rate": 6.241460580433411e-06,
+      "loss": 0.4763,
+      "step": 8172
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2005624605300265,
+      "learning_rate": 6.239132259790248e-06,
+      "loss": 0.5166,
+      "step": 8173
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1994956625652882,
+      "learning_rate": 6.23680417659007e-06,
+      "loss": 0.555,
+      "step": 8174
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1545582844827547,
+      "learning_rate": 6.234476330979859e-06,
+      "loss": 0.5358,
+      "step": 8175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1117929247561533,
+      "learning_rate": 6.232148723106586e-06,
+      "loss": 0.517,
+      "step": 8176
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1711607218520421,
+      "learning_rate": 6.2298213531172055e-06,
+      "loss": 0.5025,
+      "step": 8177
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1826246971692047,
+      "learning_rate": 6.227494221158652e-06,
+      "loss": 0.4513,
+      "step": 8178
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1192390971109658,
+      "learning_rate": 6.225167327377852e-06,
+      "loss": 0.5284,
+      "step": 8179
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.165086027415438,
+      "learning_rate": 6.222840671921715e-06,
+      "loss": 0.472,
+      "step": 8180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1796114562327191,
+      "learning_rate": 6.220514254937131e-06,
+      "loss": 0.497,
+      "step": 8181
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2233615096500101,
+      "learning_rate": 6.218188076570988e-06,
+      "loss": 0.5247,
+      "step": 8182
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1942694108202858,
+      "learning_rate": 6.215862136970139e-06,
+      "loss": 0.5362,
+      "step": 8183
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2695039716871048,
+      "learning_rate": 6.213536436281438e-06,
+      "loss": 0.573,
+      "step": 8184
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2133352843642704,
+      "learning_rate": 6.211210974651716e-06,
+      "loss": 0.4941,
+      "step": 8185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2585189446655267,
+      "learning_rate": 6.208885752227791e-06,
+      "loss": 0.5347,
+      "step": 8186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1384544623025161,
+      "learning_rate": 6.2065607691564736e-06,
+      "loss": 0.4804,
+      "step": 8187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.228707060582671,
+      "learning_rate": 6.204236025584542e-06,
+      "loss": 0.5652,
+      "step": 8188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2145456027958572,
+      "learning_rate": 6.201911521658777e-06,
+      "loss": 0.4816,
+      "step": 8189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1635301127780793,
+      "learning_rate": 6.19958725752593e-06,
+      "loss": 0.541,
+      "step": 8190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1273108696322753,
+      "learning_rate": 6.197263233332747e-06,
+      "loss": 0.5061,
+      "step": 8191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1827566378061876,
+      "learning_rate": 6.19493944922596e-06,
+      "loss": 0.5096,
+      "step": 8192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2844155910788315,
+      "learning_rate": 6.192615905352273e-06,
+      "loss": 0.521,
+      "step": 8193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1024948395069822,
+      "learning_rate": 6.190292601858389e-06,
+      "loss": 0.4921,
+      "step": 8194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1057808371480182,
+      "learning_rate": 6.1879695388909865e-06,
+      "loss": 0.441,
+      "step": 8195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2051472698140704,
+      "learning_rate": 6.185646716596735e-06,
+      "loss": 0.4912,
+      "step": 8196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.262531408542438,
+      "learning_rate": 6.183324135122289e-06,
+      "loss": 0.5507,
+      "step": 8197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2005207561015714,
+      "learning_rate": 6.181001794614279e-06,
+      "loss": 0.5171,
+      "step": 8198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2581858584318575,
+      "learning_rate": 6.17867969521933e-06,
+      "loss": 0.5506,
+      "step": 8199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0402499419425246,
+      "learning_rate": 6.176357837084046e-06,
+      "loss": 0.5074,
+      "step": 8200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.339137789413524,
+      "learning_rate": 6.17403622035502e-06,
+      "loss": 0.5281,
+      "step": 8201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2314929877301088,
+      "learning_rate": 6.1717148451788265e-06,
+      "loss": 0.5521,
+      "step": 8202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2629923814013706,
+      "learning_rate": 6.169393711702027e-06,
+      "loss": 0.5692,
+      "step": 8203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1138745784765145,
+      "learning_rate": 6.167072820071167e-06,
+      "loss": 0.5313,
+      "step": 8204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1767591294403938,
+      "learning_rate": 6.164752170432773e-06,
+      "loss": 0.5292,
+      "step": 8205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1553288476317225,
+      "learning_rate": 6.162431762933361e-06,
+      "loss": 0.531,
+      "step": 8206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1536558795201628,
+      "learning_rate": 6.160111597719433e-06,
+      "loss": 0.5361,
+      "step": 8207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3086055470404578,
+      "learning_rate": 6.157791674937471e-06,
+      "loss": 0.5402,
+      "step": 8208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2474205582729367,
+      "learning_rate": 6.1554719947339435e-06,
+      "loss": 0.5066,
+      "step": 8209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1854479275533065,
+      "learning_rate": 6.153152557255303e-06,
+      "loss": 0.5158,
+      "step": 8210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2329209854103622,
+      "learning_rate": 6.150833362647988e-06,
+      "loss": 0.5584,
+      "step": 8211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3002949856818755,
+      "learning_rate": 6.148514411058424e-06,
+      "loss": 0.5073,
+      "step": 8212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1295832415704656,
+      "learning_rate": 6.146195702633018e-06,
+      "loss": 0.5404,
+      "step": 8213
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1528537462217887,
+      "learning_rate": 6.143877237518158e-06,
+      "loss": 0.562,
+      "step": 8214
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2224765925906589,
+      "learning_rate": 6.141559015860221e-06,
+      "loss": 0.5699,
+      "step": 8215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1697335067698311,
+      "learning_rate": 6.1392410378055725e-06,
+      "loss": 0.5119,
+      "step": 8216
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2611346236084933,
+      "learning_rate": 6.136923303500556e-06,
+      "loss": 0.5468,
+      "step": 8217
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3625432068859702,
+      "learning_rate": 6.134605813091503e-06,
+      "loss": 0.5605,
+      "step": 8218
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0981414374226928,
+      "learning_rate": 6.132288566724728e-06,
+      "loss": 0.5418,
+      "step": 8219
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1336824956815332,
+      "learning_rate": 6.129971564546529e-06,
+      "loss": 0.4852,
+      "step": 8220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0734901315371415,
+      "learning_rate": 6.127654806703189e-06,
+      "loss": 0.493,
+      "step": 8221
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1775994421239755,
+      "learning_rate": 6.125338293340985e-06,
+      "loss": 0.5664,
+      "step": 8222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0850686089406505,
+      "learning_rate": 6.123022024606165e-06,
+      "loss": 0.5001,
+      "step": 8223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.152421334451245,
+      "learning_rate": 6.120706000644965e-06,
+      "loss": 0.5164,
+      "step": 8224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1588395992892193,
+      "learning_rate": 6.11839022160361e-06,
+      "loss": 0.5319,
+      "step": 8225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2549046139734341,
+      "learning_rate": 6.116074687628305e-06,
+      "loss": 0.5439,
+      "step": 8226
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.197136279239915,
+      "learning_rate": 6.113759398865247e-06,
+      "loss": 0.5232,
+      "step": 8227
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2277228825486712,
+      "learning_rate": 6.111444355460608e-06,
+      "loss": 0.512,
+      "step": 8228
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1986172438094187,
+      "learning_rate": 6.109129557560547e-06,
+      "loss": 0.5357,
+      "step": 8229
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2094532391772521,
+      "learning_rate": 6.106815005311211e-06,
+      "loss": 0.5506,
+      "step": 8230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3012841887659479,
+      "learning_rate": 6.104500698858731e-06,
+      "loss": 0.5308,
+      "step": 8231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.177776835674133,
+      "learning_rate": 6.1021866383492205e-06,
+      "loss": 0.5641,
+      "step": 8232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1547059755821678,
+      "learning_rate": 6.0998728239287784e-06,
+      "loss": 0.4707,
+      "step": 8233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1894349846920798,
+      "learning_rate": 6.097559255743486e-06,
+      "loss": 0.5333,
+      "step": 8234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.167801861019125,
+      "learning_rate": 6.095245933939411e-06,
+      "loss": 0.5321,
+      "step": 8235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2140268111905477,
+      "learning_rate": 6.092932858662604e-06,
+      "loss": 0.5507,
+      "step": 8236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.243314124172889,
+      "learning_rate": 6.0906200300591074e-06,
+      "loss": 0.5647,
+      "step": 8237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0433527207528352,
+      "learning_rate": 6.088307448274937e-06,
+      "loss": 0.4527,
+      "step": 8238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2909960954490376,
+      "learning_rate": 6.0859951134561e-06,
+      "loss": 0.5848,
+      "step": 8239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0789159554743135,
+      "learning_rate": 6.083683025748584e-06,
+      "loss": 0.4831,
+      "step": 8240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2012878858167118,
+      "learning_rate": 6.081371185298361e-06,
+      "loss": 0.4837,
+      "step": 8241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.304273345451697,
+      "learning_rate": 6.079059592251398e-06,
+      "loss": 0.5552,
+      "step": 8242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1499823216447809,
+      "learning_rate": 6.076748246753632e-06,
+      "loss": 0.5411,
+      "step": 8243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.202401030070231,
+      "learning_rate": 6.074437148950987e-06,
+      "loss": 0.5333,
+      "step": 8244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1568132523555825,
+      "learning_rate": 6.072126298989378e-06,
+      "loss": 0.5068,
+      "step": 8245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0978312505591579,
+      "learning_rate": 6.069815697014701e-06,
+      "loss": 0.4789,
+      "step": 8246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1675491344723636,
+      "learning_rate": 6.067505343172839e-06,
+      "loss": 0.5373,
+      "step": 8247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2823501608692853,
+      "learning_rate": 6.065195237609655e-06,
+      "loss": 0.5246,
+      "step": 8248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2692917818019045,
+      "learning_rate": 6.062885380470992e-06,
+      "loss": 0.524,
+      "step": 8249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2624052559833838,
+      "learning_rate": 6.0605757719026884e-06,
+      "loss": 0.5993,
+      "step": 8250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1384870794716562,
+      "learning_rate": 6.058266412050561e-06,
+      "loss": 0.5219,
+      "step": 8251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1366436595956209,
+      "learning_rate": 6.055957301060413e-06,
+      "loss": 0.5149,
+      "step": 8252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1804160085492914,
+      "learning_rate": 6.053648439078033e-06,
+      "loss": 0.5218,
+      "step": 8253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1742472733353682,
+      "learning_rate": 6.05133982624918e-06,
+      "loss": 0.5603,
+      "step": 8254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1696696065175627,
+      "learning_rate": 6.04903146271962e-06,
+      "loss": 0.5343,
+      "step": 8255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0823302148606238,
+      "learning_rate": 6.046723348635086e-06,
+      "loss": 0.5017,
+      "step": 8256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7344631825616788,
+      "learning_rate": 6.044415484141306e-06,
+      "loss": 0.5411,
+      "step": 8257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2492929843319336,
+      "learning_rate": 6.042107869383982e-06,
+      "loss": 0.535,
+      "step": 8258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2330886318839456,
+      "learning_rate": 6.039800504508813e-06,
+      "loss": 0.504,
+      "step": 8259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1370313286738691,
+      "learning_rate": 6.0374933896614665e-06,
+      "loss": 0.5477,
+      "step": 8260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.235705515698876,
+      "learning_rate": 6.035186524987605e-06,
+      "loss": 0.5588,
+      "step": 8261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2104577867851762,
+      "learning_rate": 6.032879910632876e-06,
+      "loss": 0.5505,
+      "step": 8262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1999942123750353,
+      "learning_rate": 6.030573546742904e-06,
+      "loss": 0.5093,
+      "step": 8263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2554652899437484,
+      "learning_rate": 6.028267433463309e-06,
+      "loss": 0.5354,
+      "step": 8264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1396285106809467,
+      "learning_rate": 6.025961570939676e-06,
+      "loss": 0.482,
+      "step": 8265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2477806893948147,
+      "learning_rate": 6.023655959317594e-06,
+      "loss": 0.5834,
+      "step": 8266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.178961776932395,
+      "learning_rate": 6.021350598742628e-06,
+      "loss": 0.5392,
+      "step": 8267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2412358603387617,
+      "learning_rate": 6.019045489360325e-06,
+      "loss": 0.5755,
+      "step": 8268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1502549718074888,
+      "learning_rate": 6.016740631316221e-06,
+      "loss": 0.5257,
+      "step": 8269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2423343691581445,
+      "learning_rate": 6.01443602475583e-06,
+      "loss": 0.5144,
+      "step": 8270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.192894328356021,
+      "learning_rate": 6.0121316698246535e-06,
+      "loss": 0.5078,
+      "step": 8271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2212386032556652,
+      "learning_rate": 6.009827566668183e-06,
+      "loss": 0.5382,
+      "step": 8272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1543874947078763,
+      "learning_rate": 6.007523715431882e-06,
+      "loss": 0.5206,
+      "step": 8273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2207255369036356,
+      "learning_rate": 6.0052201162612125e-06,
+      "loss": 0.5538,
+      "step": 8274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0982657807048068,
+      "learning_rate": 6.002916769301601e-06,
+      "loss": 0.4986,
+      "step": 8275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2048283693748971,
+      "learning_rate": 6.000613674698478e-06,
+      "loss": 0.5504,
+      "step": 8276
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1398957939372207,
+      "learning_rate": 5.99831083259725e-06,
+      "loss": 0.497,
+      "step": 8277
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.192742370830532,
+      "learning_rate": 5.996008243143302e-06,
+      "loss": 0.5538,
+      "step": 8278
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1809412376100377,
+      "learning_rate": 5.993705906482016e-06,
+      "loss": 0.483,
+      "step": 8279
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2401758375395675,
+      "learning_rate": 5.991403822758741e-06,
+      "loss": 0.5735,
+      "step": 8280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2014972029108515,
+      "learning_rate": 5.9891019921188264e-06,
+      "loss": 0.5472,
+      "step": 8281
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1871760076944107,
+      "learning_rate": 5.986800414707596e-06,
+      "loss": 0.5072,
+      "step": 8282
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.065242873333841,
+      "learning_rate": 5.984499090670361e-06,
+      "loss": 0.4722,
+      "step": 8283
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2389210392943288,
+      "learning_rate": 5.982198020152419e-06,
+      "loss": 0.4703,
+      "step": 8284
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2349325441942252,
+      "learning_rate": 5.979897203299041e-06,
+      "loss": 0.5288,
+      "step": 8285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1463681215681572,
+      "learning_rate": 5.977596640255494e-06,
+      "loss": 0.5446,
+      "step": 8286
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.200546275411955,
+      "learning_rate": 5.975296331167025e-06,
+      "loss": 0.5458,
+      "step": 8287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.229241719082832,
+      "learning_rate": 5.972996276178862e-06,
+      "loss": 0.5615,
+      "step": 8288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0795875801190369,
+      "learning_rate": 5.970696475436224e-06,
+      "loss": 0.5118,
+      "step": 8289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1201241051510997,
+      "learning_rate": 5.968396929084303e-06,
+      "loss": 0.5111,
+      "step": 8290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.21812701808866,
+      "learning_rate": 5.966097637268284e-06,
+      "loss": 0.5223,
+      "step": 8291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1651956048564998,
+      "learning_rate": 5.963798600133334e-06,
+      "loss": 0.5376,
+      "step": 8292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0966688174272066,
+      "learning_rate": 5.961499817824603e-06,
+      "loss": 0.4903,
+      "step": 8293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1281372407793313,
+      "learning_rate": 5.959201290487227e-06,
+      "loss": 0.5058,
+      "step": 8294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0451609115279918,
+      "learning_rate": 5.956903018266317e-06,
+      "loss": 0.46,
+      "step": 8295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.02968762271863,
+      "learning_rate": 5.954605001306979e-06,
+      "loss": 0.4942,
+      "step": 8296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1389752829484674,
+      "learning_rate": 5.952307239754302e-06,
+      "loss": 0.5019,
+      "step": 8297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0504699018870116,
+      "learning_rate": 5.950009733753348e-06,
+      "loss": 0.505,
+      "step": 8298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2733212751463376,
+      "learning_rate": 5.94771248344918e-06,
+      "loss": 0.6072,
+      "step": 8299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1068717222418305,
+      "learning_rate": 5.945415488986827e-06,
+      "loss": 0.5147,
+      "step": 8300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.140142573927685,
+      "learning_rate": 5.943118750511312e-06,
+      "loss": 0.5457,
+      "step": 8301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1281983687194435,
+      "learning_rate": 5.940822268167643e-06,
+      "loss": 0.5293,
+      "step": 8302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1480746052687518,
+      "learning_rate": 5.938526042100805e-06,
+      "loss": 0.5418,
+      "step": 8303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0439502227068225,
+      "learning_rate": 5.936230072455777e-06,
+      "loss": 0.5033,
+      "step": 8304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1160185240059313,
+      "learning_rate": 5.933934359377506e-06,
+      "loss": 0.5487,
+      "step": 8305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.130777095772815,
+      "learning_rate": 5.931638903010936e-06,
+      "loss": 0.489,
+      "step": 8306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2304808116503323,
+      "learning_rate": 5.929343703500996e-06,
+      "loss": 0.6044,
+      "step": 8307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2536175356836792,
+      "learning_rate": 5.927048760992589e-06,
+      "loss": 0.5763,
+      "step": 8308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3269634328401938,
+      "learning_rate": 5.9247540756306075e-06,
+      "loss": 0.5933,
+      "step": 8309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0907472536488314,
+      "learning_rate": 5.922459647559926e-06,
+      "loss": 0.5136,
+      "step": 8310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2105952609874586,
+      "learning_rate": 5.920165476925402e-06,
+      "loss": 0.5953,
+      "step": 8311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2333478879499373,
+      "learning_rate": 5.917871563871884e-06,
+      "loss": 0.5153,
+      "step": 8312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1397103077912958,
+      "learning_rate": 5.915577908544194e-06,
+      "loss": 0.5051,
+      "step": 8313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2077586627129169,
+      "learning_rate": 5.913284511087142e-06,
+      "loss": 0.5548,
+      "step": 8314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1768030941430447,
+      "learning_rate": 5.910991371645527e-06,
+      "loss": 0.5132,
+      "step": 8315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.077438135694472,
+      "learning_rate": 5.90869849036412e-06,
+      "loss": 0.4878,
+      "step": 8316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2826787379381503,
+      "learning_rate": 5.906405867387688e-06,
+      "loss": 0.5751,
+      "step": 8317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2665143126331475,
+      "learning_rate": 5.904113502860971e-06,
+      "loss": 0.5401,
+      "step": 8318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1841015623389202,
+      "learning_rate": 5.901821396928702e-06,
+      "loss": 0.5724,
+      "step": 8319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1058476206573513,
+      "learning_rate": 5.899529549735594e-06,
+      "loss": 0.4697,
+      "step": 8320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2409869456376905,
+      "learning_rate": 5.897237961426339e-06,
+      "loss": 0.5898,
+      "step": 8321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2756947363669284,
+      "learning_rate": 5.894946632145619e-06,
+      "loss": 0.5793,
+      "step": 8322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1509343207981138,
+      "learning_rate": 5.892655562038098e-06,
+      "loss": 0.5017,
+      "step": 8323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2297489543521822,
+      "learning_rate": 5.8903647512484205e-06,
+      "loss": 0.5716,
+      "step": 8324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2654204497638248,
+      "learning_rate": 5.888074199921223e-06,
+      "loss": 0.5429,
+      "step": 8325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.17487368614918,
+      "learning_rate": 5.885783908201114e-06,
+      "loss": 0.5171,
+      "step": 8326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1666095186042373,
+      "learning_rate": 5.883493876232693e-06,
+      "loss": 0.5151,
+      "step": 8327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2161169218864893,
+      "learning_rate": 5.8812041041605426e-06,
+      "loss": 0.5609,
+      "step": 8328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1686488184250539,
+      "learning_rate": 5.878914592129226e-06,
+      "loss": 0.5094,
+      "step": 8329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1911274067741016,
+      "learning_rate": 5.876625340283296e-06,
+      "loss": 0.5063,
+      "step": 8330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1299172596473543,
+      "learning_rate": 5.87433634876728e-06,
+      "loss": 0.501,
+      "step": 8331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2161870074058212,
+      "learning_rate": 5.872047617725697e-06,
+      "loss": 0.5331,
+      "step": 8332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0527753662734873,
+      "learning_rate": 5.869759147303042e-06,
+      "loss": 0.4686,
+      "step": 8333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2055201284024752,
+      "learning_rate": 5.867470937643804e-06,
+      "loss": 0.4812,
+      "step": 8334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1050213558843327,
+      "learning_rate": 5.865182988892449e-06,
+      "loss": 0.5358,
+      "step": 8335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1816902087188583,
+      "learning_rate": 5.862895301193421e-06,
+      "loss": 0.5774,
+      "step": 8336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.118904128725864,
+      "learning_rate": 5.86060787469116e-06,
+      "loss": 0.4766,
+      "step": 8337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1479559167612174,
+      "learning_rate": 5.858320709530077e-06,
+      "loss": 0.5263,
+      "step": 8338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.284340364680649,
+      "learning_rate": 5.8560338058545775e-06,
+      "loss": 0.5256,
+      "step": 8339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.149892909827278,
+      "learning_rate": 5.853747163809047e-06,
+      "loss": 0.5303,
+      "step": 8340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0886447559560994,
+      "learning_rate": 5.851460783537848e-06,
+      "loss": 0.5177,
+      "step": 8341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2905010173073816,
+      "learning_rate": 5.8491746651853305e-06,
+      "loss": 0.5882,
+      "step": 8342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1416802424066883,
+      "learning_rate": 5.846888808895833e-06,
+      "loss": 0.4971,
+      "step": 8343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2945659257553124,
+      "learning_rate": 5.8446032148136725e-06,
+      "loss": 0.5462,
+      "step": 8344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1650222303864493,
+      "learning_rate": 5.842317883083153e-06,
+      "loss": 0.4941,
+      "step": 8345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2245997378321456,
+      "learning_rate": 5.840032813848555e-06,
+      "loss": 0.516,
+      "step": 8346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.061853155817389,
+      "learning_rate": 5.837748007254146e-06,
+      "loss": 0.508,
+      "step": 8347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1234121242835344,
+      "learning_rate": 5.835463463444179e-06,
+      "loss": 0.5253,
+      "step": 8348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0684471303545413,
+      "learning_rate": 5.833179182562891e-06,
+      "loss": 0.4648,
+      "step": 8349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.184203289941239,
+      "learning_rate": 5.830895164754502e-06,
+      "loss": 0.5707,
+      "step": 8350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.118113900604512,
+      "learning_rate": 5.828611410163207e-06,
+      "loss": 0.453,
+      "step": 8351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.121939416592712,
+      "learning_rate": 5.826327918933197e-06,
+      "loss": 0.5192,
+      "step": 8352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.119115592849685,
+      "learning_rate": 5.824044691208641e-06,
+      "loss": 0.5349,
+      "step": 8353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1414705565168812,
+      "learning_rate": 5.821761727133686e-06,
+      "loss": 0.495,
+      "step": 8354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2759335167136183,
+      "learning_rate": 5.8194790268524725e-06,
+      "loss": 0.5713,
+      "step": 8355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2192126643033732,
+      "learning_rate": 5.817196590509113e-06,
+      "loss": 0.5445,
+      "step": 8356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4275029069166099,
+      "learning_rate": 5.814914418247714e-06,
+      "loss": 0.5057,
+      "step": 8357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2192946464150107,
+      "learning_rate": 5.812632510212359e-06,
+      "loss": 0.548,
+      "step": 8358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.169197329424752,
+      "learning_rate": 5.810350866547119e-06,
+      "loss": 0.5311,
+      "step": 8359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1895552968709187,
+      "learning_rate": 5.808069487396048e-06,
+      "loss": 0.4849,
+      "step": 8360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0370600054640842,
+      "learning_rate": 5.805788372903174e-06,
+      "loss": 0.496,
+      "step": 8361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.133862607215138,
+      "learning_rate": 5.80350752321252e-06,
+      "loss": 0.5086,
+      "step": 8362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1106481634074084,
+      "learning_rate": 5.801226938468089e-06,
+      "loss": 0.4932,
+      "step": 8363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.128475331359232,
+      "learning_rate": 5.798946618813861e-06,
+      "loss": 0.4842,
+      "step": 8364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1648247290631113,
+      "learning_rate": 5.796666564393811e-06,
+      "loss": 0.5588,
+      "step": 8365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1704829785764537,
+      "learning_rate": 5.7943867753518845e-06,
+      "loss": 0.5009,
+      "step": 8366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0497571073883374,
+      "learning_rate": 5.792107251832018e-06,
+      "loss": 0.4907,
+      "step": 8367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.202705756129117,
+      "learning_rate": 5.789827993978131e-06,
+      "loss": 0.5751,
+      "step": 8368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2167405900750052,
+      "learning_rate": 5.787549001934125e-06,
+      "loss": 0.5303,
+      "step": 8369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.196261658646584,
+      "learning_rate": 5.785270275843883e-06,
+      "loss": 0.5101,
+      "step": 8370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.207802436658382,
+      "learning_rate": 5.7829918158512774e-06,
+      "loss": 0.5534,
+      "step": 8371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1272561444394045,
+      "learning_rate": 5.7807136221001515e-06,
+      "loss": 0.5044,
+      "step": 8372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1363927104871914,
+      "learning_rate": 5.778435694734348e-06,
+      "loss": 0.5429,
+      "step": 8373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1579923647067174,
+      "learning_rate": 5.776158033897674e-06,
+      "loss": 0.5115,
+      "step": 8374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3101519516912479,
+      "learning_rate": 5.773880639733938e-06,
+      "loss": 0.623,
+      "step": 8375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2101153135917795,
+      "learning_rate": 5.771603512386923e-06,
+      "loss": 0.535,
+      "step": 8376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0191385967311948,
+      "learning_rate": 5.769326652000391e-06,
+      "loss": 0.4757,
+      "step": 8377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.160298707994621,
+      "learning_rate": 5.7670500587180935e-06,
+      "loss": 0.5258,
+      "step": 8378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0811363403028997,
+      "learning_rate": 5.764773732683766e-06,
+      "loss": 0.5088,
+      "step": 8379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2227004655636986,
+      "learning_rate": 5.7624976740411244e-06,
+      "loss": 0.5118,
+      "step": 8380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.195517079916166,
+      "learning_rate": 5.76022188293387e-06,
+      "loss": 0.5365,
+      "step": 8381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1263613940665935,
+      "learning_rate": 5.757946359505679e-06,
+      "loss": 0.5192,
+      "step": 8382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2229793231429569,
+      "learning_rate": 5.755671103900225e-06,
+      "loss": 0.5266,
+      "step": 8383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.245085592430017,
+      "learning_rate": 5.753396116261148e-06,
+      "loss": 0.5082,
+      "step": 8384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1785393995416413,
+      "learning_rate": 5.751121396732082e-06,
+      "loss": 0.4906,
+      "step": 8385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2330567769407892,
+      "learning_rate": 5.7488469454566484e-06,
+      "loss": 0.5328,
+      "step": 8386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.175403217363587,
+      "learning_rate": 5.746572762578437e-06,
+      "loss": 0.5163,
+      "step": 8387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.206452956444843,
+      "learning_rate": 5.744298848241032e-06,
+      "loss": 0.5356,
+      "step": 8388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2145841265894468,
+      "learning_rate": 5.742025202587997e-06,
+      "loss": 0.5426,
+      "step": 8389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2460494078396926,
+      "learning_rate": 5.739751825762878e-06,
+      "loss": 0.6032,
+      "step": 8390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0768910409149992,
+      "learning_rate": 5.7374787179092106e-06,
+      "loss": 0.5589,
+      "step": 8391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1603555732867057,
+      "learning_rate": 5.7352058791705e-06,
+      "loss": 0.5156,
+      "step": 8392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1952020930742986,
+      "learning_rate": 5.732933309690251e-06,
+      "loss": 0.5132,
+      "step": 8393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2150487696117778,
+      "learning_rate": 5.730661009611931e-06,
+      "loss": 0.5564,
+      "step": 8394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0558316841403654,
+      "learning_rate": 5.7283889790790096e-06,
+      "loss": 0.4844,
+      "step": 8395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1781275458270426,
+      "learning_rate": 5.7261172182349344e-06,
+      "loss": 0.574,
+      "step": 8396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2799412160279555,
+      "learning_rate": 5.723845727223125e-06,
+      "loss": 0.5643,
+      "step": 8397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.057899726190161,
+      "learning_rate": 5.721574506186998e-06,
+      "loss": 0.4961,
+      "step": 8398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0321424264100818,
+      "learning_rate": 5.719303555269946e-06,
+      "loss": 0.4315,
+      "step": 8399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1565343146405827,
+      "learning_rate": 5.717032874615345e-06,
+      "loss": 0.5138,
+      "step": 8400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1410248460563566,
+      "learning_rate": 5.714762464366561e-06,
+      "loss": 0.4653,
+      "step": 8401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.128930220299961,
+      "learning_rate": 5.712492324666927e-06,
+      "loss": 0.5215,
+      "step": 8402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2231399992898826,
+      "learning_rate": 5.7102224556597775e-06,
+      "loss": 0.5664,
+      "step": 8403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2105101292693339,
+      "learning_rate": 5.7079528574884125e-06,
+      "loss": 0.5789,
+      "step": 8404
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1467437710639632,
+      "learning_rate": 5.7056835302961266e-06,
+      "loss": 0.4865,
+      "step": 8405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1429240950063595,
+      "learning_rate": 5.703414474226201e-06,
+      "loss": 0.5174,
+      "step": 8406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1908314303166476,
+      "learning_rate": 5.701145689421882e-06,
+      "loss": 0.5352,
+      "step": 8407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2017686307418005,
+      "learning_rate": 5.698877176026415e-06,
+      "loss": 0.5329,
+      "step": 8408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.210155012747007,
+      "learning_rate": 5.696608934183023e-06,
+      "loss": 0.5465,
+      "step": 8409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.186342227112146,
+      "learning_rate": 5.694340964034911e-06,
+      "loss": 0.5734,
+      "step": 8410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1401453969490392,
+      "learning_rate": 5.692073265725273e-06,
+      "loss": 0.484,
+      "step": 8411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0952495105170241,
+      "learning_rate": 5.689805839397271e-06,
+      "loss": 0.5017,
+      "step": 8412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2199979896997868,
+      "learning_rate": 5.687538685194069e-06,
+      "loss": 0.4978,
+      "step": 8413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.077079431808361,
+      "learning_rate": 5.685271803258794e-06,
+      "loss": 0.4802,
+      "step": 8414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2055374333610978,
+      "learning_rate": 5.683005193734572e-06,
+      "loss": 0.5326,
+      "step": 8415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1562690733290126,
+      "learning_rate": 5.680738856764508e-06,
+      "loss": 0.5369,
+      "step": 8416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1738617395548328,
+      "learning_rate": 5.67847279249168e-06,
+      "loss": 0.5227,
+      "step": 8417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2168957224660126,
+      "learning_rate": 5.676207001059163e-06,
+      "loss": 0.5432,
+      "step": 8418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0244021346831507,
+      "learning_rate": 5.673941482610004e-06,
+      "loss": 0.4651,
+      "step": 8419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1248860301462438,
+      "learning_rate": 5.67167623728724e-06,
+      "loss": 0.4723,
+      "step": 8420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0739386175931611,
+      "learning_rate": 5.6694112652338895e-06,
+      "loss": 0.5137,
+      "step": 8421
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.059763301027072,
+      "learning_rate": 5.667146566592945e-06,
+      "loss": 0.4813,
+      "step": 8422
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.20870554332382,
+      "learning_rate": 5.6648821415073965e-06,
+      "loss": 0.49,
+      "step": 8423
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0993197179864558,
+      "learning_rate": 5.662617990120201e-06,
+      "loss": 0.4769,
+      "step": 8424
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.270928749296342,
+      "learning_rate": 5.660354112574309e-06,
+      "loss": 0.5251,
+      "step": 8425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.147109373087845,
+      "learning_rate": 5.658090509012651e-06,
+      "loss": 0.5149,
+      "step": 8426
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2236796717945462,
+      "learning_rate": 5.655827179578145e-06,
+      "loss": 0.5439,
+      "step": 8427
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1690986296927226,
+      "learning_rate": 5.653564124413678e-06,
+      "loss": 0.4971,
+      "step": 8428
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1824483837612974,
+      "learning_rate": 5.651301343662132e-06,
+      "loss": 0.5388,
+      "step": 8429
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0618730827255858,
+      "learning_rate": 5.649038837466369e-06,
+      "loss": 0.4805,
+      "step": 8430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1365684069022948,
+      "learning_rate": 5.646776605969237e-06,
+      "loss": 0.5238,
+      "step": 8431
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1641473931201003,
+      "learning_rate": 5.644514649313554e-06,
+      "loss": 0.5511,
+      "step": 8432
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2852973744554859,
+      "learning_rate": 5.642252967642134e-06,
+      "loss": 0.531,
+      "step": 8433
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2137818433078988,
+      "learning_rate": 5.639991561097767e-06,
+      "loss": 0.5753,
+      "step": 8434
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1422237009555516,
+      "learning_rate": 5.637730429823224e-06,
+      "loss": 0.5511,
+      "step": 8435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1509193021654671,
+      "learning_rate": 5.6354695739612665e-06,
+      "loss": 0.4956,
+      "step": 8436
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1055873631982116,
+      "learning_rate": 5.6332089936546375e-06,
+      "loss": 0.5111,
+      "step": 8437
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1053391774942984,
+      "learning_rate": 5.6309486890460494e-06,
+      "loss": 0.5161,
+      "step": 8438
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1290814748817206,
+      "learning_rate": 5.628688660278212e-06,
+      "loss": 0.5155,
+      "step": 8439
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1467821816452228,
+      "learning_rate": 5.6264289074938126e-06,
+      "loss": 0.5373,
+      "step": 8440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.32883142869552,
+      "learning_rate": 5.624169430835524e-06,
+      "loss": 0.5678,
+      "step": 8441
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2227426809163227,
+      "learning_rate": 5.621910230445993e-06,
+      "loss": 0.5664,
+      "step": 8442
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2090601982041547,
+      "learning_rate": 5.619651306467861e-06,
+      "loss": 0.5866,
+      "step": 8443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.195964809770765,
+      "learning_rate": 5.617392659043737e-06,
+      "loss": 0.4943,
+      "step": 8444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.077583453405085,
+      "learning_rate": 5.615134288316227e-06,
+      "loss": 0.551,
+      "step": 8445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1009499761008235,
+      "learning_rate": 5.612876194427911e-06,
+      "loss": 0.526,
+      "step": 8446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0947278692287028,
+      "learning_rate": 5.61061837752136e-06,
+      "loss": 0.489,
+      "step": 8447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1213544482655873,
+      "learning_rate": 5.608360837739113e-06,
+      "loss": 0.5151,
+      "step": 8448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2054608941208498,
+      "learning_rate": 5.6061035752237035e-06,
+      "loss": 0.4951,
+      "step": 8449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1468636764258395,
+      "learning_rate": 5.6038465901176455e-06,
+      "loss": 0.5374,
+      "step": 8450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1641627530956273,
+      "learning_rate": 5.601589882563436e-06,
+      "loss": 0.493,
+      "step": 8451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1911011852486217,
+      "learning_rate": 5.599333452703548e-06,
+      "loss": 0.5355,
+      "step": 8452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1434821831499857,
+      "learning_rate": 5.5970773006804465e-06,
+      "loss": 0.5303,
+      "step": 8453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1601964256445938,
+      "learning_rate": 5.594821426636567e-06,
+      "loss": 0.5065,
+      "step": 8454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0777001580464676,
+      "learning_rate": 5.5925658307143405e-06,
+      "loss": 0.4361,
+      "step": 8455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1149810663677027,
+      "learning_rate": 5.590310513056171e-06,
+      "loss": 0.5034,
+      "step": 8456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0965889734402938,
+      "learning_rate": 5.588055473804453e-06,
+      "loss": 0.5395,
+      "step": 8457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2157905054077165,
+      "learning_rate": 5.585800713101552e-06,
+      "loss": 0.5396,
+      "step": 8458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1496163370930719,
+      "learning_rate": 5.583546231089827e-06,
+      "loss": 0.5204,
+      "step": 8459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1547193447981894,
+      "learning_rate": 5.581292027911614e-06,
+      "loss": 0.5335,
+      "step": 8460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1413997475023059,
+      "learning_rate": 5.579038103709238e-06,
+      "loss": 0.5011,
+      "step": 8461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1848394806704596,
+      "learning_rate": 5.576784458624991e-06,
+      "loss": 0.5875,
+      "step": 8462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1876907195479356,
+      "learning_rate": 5.5745310928011656e-06,
+      "loss": 0.5729,
+      "step": 8463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2527726893428552,
+      "learning_rate": 5.57227800638002e-06,
+      "loss": 0.5458,
+      "step": 8464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1716997651371692,
+      "learning_rate": 5.570025199503808e-06,
+      "loss": 0.5917,
+      "step": 8465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1851059220855302,
+      "learning_rate": 5.567772672314762e-06,
+      "loss": 0.5326,
+      "step": 8466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1172082538944497,
+      "learning_rate": 5.565520424955097e-06,
+      "loss": 0.4709,
+      "step": 8467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2249893985990217,
+      "learning_rate": 5.563268457567004e-06,
+      "loss": 0.5497,
+      "step": 8468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1785875459700161,
+      "learning_rate": 5.561016770292662e-06,
+      "loss": 0.5436,
+      "step": 8469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2230458476751565,
+      "learning_rate": 5.558765363274234e-06,
+      "loss": 0.5536,
+      "step": 8470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1308483063941799,
+      "learning_rate": 5.556514236653867e-06,
+      "loss": 0.5109,
+      "step": 8471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1662427211423936,
+      "learning_rate": 5.5542633905736775e-06,
+      "loss": 0.5295,
+      "step": 8472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1748412674356221,
+      "learning_rate": 5.552012825175781e-06,
+      "loss": 0.5335,
+      "step": 8473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2096217729218548,
+      "learning_rate": 5.549762540602261e-06,
+      "loss": 0.5262,
+      "step": 8474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2318628073852937,
+      "learning_rate": 5.54751253699519e-06,
+      "loss": 0.5159,
+      "step": 8475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1905554563382166,
+      "learning_rate": 5.545262814496625e-06,
+      "loss": 0.5302,
+      "step": 8476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1472164591618066,
+      "learning_rate": 5.543013373248601e-06,
+      "loss": 0.5139,
+      "step": 8477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.238370537171197,
+      "learning_rate": 5.540764213393144e-06,
+      "loss": 0.5426,
+      "step": 8478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1846252579615788,
+      "learning_rate": 5.538515335072243e-06,
+      "loss": 0.5121,
+      "step": 8479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2225057003457815,
+      "learning_rate": 5.536266738427886e-06,
+      "loss": 0.5435,
+      "step": 8480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2056883715105913,
+      "learning_rate": 5.534018423602047e-06,
+      "loss": 0.5828,
+      "step": 8481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2752135827182332,
+      "learning_rate": 5.531770390736659e-06,
+      "loss": 0.5625,
+      "step": 8482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1836965283378318,
+      "learning_rate": 5.529522639973666e-06,
+      "loss": 0.5268,
+      "step": 8483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.223927625987873,
+      "learning_rate": 5.527275171454969e-06,
+      "loss": 0.4761,
+      "step": 8484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2690409491164432,
+      "learning_rate": 5.525027985322464e-06,
+      "loss": 0.5811,
+      "step": 8485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.160096394956788,
+      "learning_rate": 5.5227810817180325e-06,
+      "loss": 0.478,
+      "step": 8486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2689591747417468,
+      "learning_rate": 5.520534460783531e-06,
+      "loss": 0.5614,
+      "step": 8487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1784244875112335,
+      "learning_rate": 5.5182881226608035e-06,
+      "loss": 0.5059,
+      "step": 8488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1366005013432052,
+      "learning_rate": 5.516042067491665e-06,
+      "loss": 0.5046,
+      "step": 8489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1004874753191654,
+      "learning_rate": 5.513796295417925e-06,
+      "loss": 0.486,
+      "step": 8490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1285325326045819,
+      "learning_rate": 5.511550806581374e-06,
+      "loss": 0.5122,
+      "step": 8491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1319279867794765,
+      "learning_rate": 5.5093056011237755e-06,
+      "loss": 0.4847,
+      "step": 8492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1975101389286498,
+      "learning_rate": 5.507060679186886e-06,
+      "loss": 0.5649,
+      "step": 8493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0671562938679564,
+      "learning_rate": 5.504816040912433e-06,
+      "loss": 0.5085,
+      "step": 8494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0458487419624567,
+      "learning_rate": 5.5025716864421356e-06,
+      "loss": 0.4772,
+      "step": 8495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.240019195392685,
+      "learning_rate": 5.500327615917691e-06,
+      "loss": 0.5743,
+      "step": 8496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2476582526547406,
+      "learning_rate": 5.498083829480778e-06,
+      "loss": 0.5587,
+      "step": 8497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0980409647138942,
+      "learning_rate": 5.495840327273065e-06,
+      "loss": 0.5549,
+      "step": 8498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0337226585453367,
+      "learning_rate": 5.493597109436186e-06,
+      "loss": 0.456,
+      "step": 8499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.113755724000618,
+      "learning_rate": 5.491354176111773e-06,
+      "loss": 0.5262,
+      "step": 8500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0530288086380135,
+      "learning_rate": 5.489111527441435e-06,
+      "loss": 0.5012,
+      "step": 8501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2221671396664937,
+      "learning_rate": 5.486869163566756e-06,
+      "loss": 0.5657,
+      "step": 8502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.070779461872384,
+      "learning_rate": 5.4846270846293145e-06,
+      "loss": 0.4805,
+      "step": 8503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.157434938880406,
+      "learning_rate": 5.4823852907706585e-06,
+      "loss": 0.4983,
+      "step": 8504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1750843606249153,
+      "learning_rate": 5.480143782132327e-06,
+      "loss": 0.5561,
+      "step": 8505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.20566795418941,
+      "learning_rate": 5.477902558855837e-06,
+      "loss": 0.5203,
+      "step": 8506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1302033466750618,
+      "learning_rate": 5.475661621082689e-06,
+      "loss": 0.5136,
+      "step": 8507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2599388299202012,
+      "learning_rate": 5.4734209689543705e-06,
+      "loss": 0.5404,
+      "step": 8508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1136368568398085,
+      "learning_rate": 5.471180602612336e-06,
+      "loss": 0.453,
+      "step": 8509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1352565314149756,
+      "learning_rate": 5.468940522198036e-06,
+      "loss": 0.4893,
+      "step": 8510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2023863568604916,
+      "learning_rate": 5.4667007278529015e-06,
+      "loss": 0.496,
+      "step": 8511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1360211411107353,
+      "learning_rate": 5.464461219718336e-06,
+      "loss": 0.4935,
+      "step": 8512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1070736399781234,
+      "learning_rate": 5.462221997935737e-06,
+      "loss": 0.4995,
+      "step": 8513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.260995616306245,
+      "learning_rate": 5.45998306264647e-06,
+      "loss": 0.4525,
+      "step": 8514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2252535791101584,
+      "learning_rate": 5.457744413991897e-06,
+      "loss": 0.5422,
+      "step": 8515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2797914902909833,
+      "learning_rate": 5.455506052113354e-06,
+      "loss": 0.5789,
+      "step": 8516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1960791329674854,
+      "learning_rate": 5.453267977152161e-06,
+      "loss": 0.5278,
+      "step": 8517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2269451734982912,
+      "learning_rate": 5.4510301892496224e-06,
+      "loss": 0.5339,
+      "step": 8518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1615682929066196,
+      "learning_rate": 5.448792688547012e-06,
+      "loss": 0.4901,
+      "step": 8519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1604961064448818,
+      "learning_rate": 5.446555475185602e-06,
+      "loss": 0.5481,
+      "step": 8520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2052080526436777,
+      "learning_rate": 5.444318549306641e-06,
+      "loss": 0.5326,
+      "step": 8521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.240185257547323,
+      "learning_rate": 5.44208191105135e-06,
+      "loss": 0.5421,
+      "step": 8522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1496226624613037,
+      "learning_rate": 5.439845560560948e-06,
+      "loss": 0.5259,
+      "step": 8523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2448385966670596,
+      "learning_rate": 5.437609497976619e-06,
+      "loss": 0.5128,
+      "step": 8524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1222563772753282,
+      "learning_rate": 5.435373723439541e-06,
+      "loss": 0.4668,
+      "step": 8525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2988610284982853,
+      "learning_rate": 5.43313823709087e-06,
+      "loss": 0.5933,
+      "step": 8526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9877893180987131,
+      "learning_rate": 5.430903039071744e-06,
+      "loss": 0.4621,
+      "step": 8527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.171360716823358,
+      "learning_rate": 5.428668129523288e-06,
+      "loss": 0.5426,
+      "step": 8528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1813154293286559,
+      "learning_rate": 5.426433508586593e-06,
+      "loss": 0.51,
+      "step": 8529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2110931788632529,
+      "learning_rate": 5.4241991764027464e-06,
+      "loss": 0.5136,
+      "step": 8530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1125414615577245,
+      "learning_rate": 5.421965133112818e-06,
+      "loss": 0.5046,
+      "step": 8531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2388047035525844,
+      "learning_rate": 5.419731378857849e-06,
+      "loss": 0.5183,
+      "step": 8532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2087526361225658,
+      "learning_rate": 5.417497913778866e-06,
+      "loss": 0.5537,
+      "step": 8533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1528444398553923,
+      "learning_rate": 5.4152647380168876e-06,
+      "loss": 0.5384,
+      "step": 8534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2111791554175395,
+      "learning_rate": 5.413031851712895e-06,
+      "loss": 0.5742,
+      "step": 8535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2373031459639174,
+      "learning_rate": 5.410799255007868e-06,
+      "loss": 0.5383,
+      "step": 8536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2333312148428242,
+      "learning_rate": 5.408566948042762e-06,
+      "loss": 0.5188,
+      "step": 8537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.103125903423685,
+      "learning_rate": 5.406334930958513e-06,
+      "loss": 0.4999,
+      "step": 8538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2777737235037205,
+      "learning_rate": 5.404103203896044e-06,
+      "loss": 0.5057,
+      "step": 8539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2015502333790002,
+      "learning_rate": 5.401871766996247e-06,
+      "loss": 0.5329,
+      "step": 8540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.23725940407713,
+      "learning_rate": 5.399640620400013e-06,
+      "loss": 0.5443,
+      "step": 8541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2897518742150835,
+      "learning_rate": 5.397409764248197e-06,
+      "loss": 0.5461,
+      "step": 8542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1129164784120786,
+      "learning_rate": 5.395179198681648e-06,
+      "loss": 0.5037,
+      "step": 8543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.149926083718583,
+      "learning_rate": 5.392948923841199e-06,
+      "loss": 0.5661,
+      "step": 8544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1224459159752647,
+      "learning_rate": 5.39071893986765e-06,
+      "loss": 0.4964,
+      "step": 8545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.277206413515417,
+      "learning_rate": 5.388489246901796e-06,
+      "loss": 0.5627,
+      "step": 8546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2132288266245856,
+      "learning_rate": 5.386259845084405e-06,
+      "loss": 0.5584,
+      "step": 8547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1921675454165848,
+      "learning_rate": 5.384030734556236e-06,
+      "loss": 0.5132,
+      "step": 8548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2253165264301842,
+      "learning_rate": 5.381801915458026e-06,
+      "loss": 0.5458,
+      "step": 8549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1537922693777483,
+      "learning_rate": 5.379573387930484e-06,
+      "loss": 0.4989,
+      "step": 8550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1037426200265723,
+      "learning_rate": 5.377345152114315e-06,
+      "loss": 0.4588,
+      "step": 8551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2108923134525391,
+      "learning_rate": 5.3751172081501935e-06,
+      "loss": 0.517,
+      "step": 8552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3200071986320028,
+      "learning_rate": 5.372889556178782e-06,
+      "loss": 0.6122,
+      "step": 8553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1628269064056198,
+      "learning_rate": 5.370662196340732e-06,
+      "loss": 0.5179,
+      "step": 8554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.228707448663035,
+      "learning_rate": 5.368435128776657e-06,
+      "loss": 0.5027,
+      "step": 8555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.142716410819025,
+      "learning_rate": 5.366208353627167e-06,
+      "loss": 0.4519,
+      "step": 8556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3166599199065854,
+      "learning_rate": 5.363981871032852e-06,
+      "loss": 0.5132,
+      "step": 8557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1404255731857622,
+      "learning_rate": 5.36175568113428e-06,
+      "loss": 0.4961,
+      "step": 8558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.214787227382845,
+      "learning_rate": 5.359529784072006e-06,
+      "loss": 0.6058,
+      "step": 8559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2205537506263417,
+      "learning_rate": 5.357304179986553e-06,
+      "loss": 0.5143,
+      "step": 8560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1882664064462138,
+      "learning_rate": 5.355078869018446e-06,
+      "loss": 0.5136,
+      "step": 8561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1278910576616812,
+      "learning_rate": 5.3528538513081705e-06,
+      "loss": 0.5075,
+      "step": 8562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.177618017822402,
+      "learning_rate": 5.350629126996207e-06,
+      "loss": 0.571,
+      "step": 8563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.219897732356841,
+      "learning_rate": 5.34840469622302e-06,
+      "loss": 0.5951,
+      "step": 8564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1129376868449574,
+      "learning_rate": 5.346180559129037e-06,
+      "loss": 0.5193,
+      "step": 8565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2290367876821564,
+      "learning_rate": 5.3439567158546865e-06,
+      "loss": 0.5916,
+      "step": 8566
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1429654500083193,
+      "learning_rate": 5.341733166540372e-06,
+      "loss": 0.5337,
+      "step": 8567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3126618875846319,
+      "learning_rate": 5.339509911326475e-06,
+      "loss": 0.5541,
+      "step": 8568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2042186274861924,
+      "learning_rate": 5.337286950353366e-06,
+      "loss": 0.5425,
+      "step": 8569
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1255786255429039,
+      "learning_rate": 5.3350642837613845e-06,
+      "loss": 0.555,
+      "step": 8570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1246557238754393,
+      "learning_rate": 5.332841911690867e-06,
+      "loss": 0.5074,
+      "step": 8571
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1832620892546886,
+      "learning_rate": 5.330619834282116e-06,
+      "loss": 0.5007,
+      "step": 8572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.231502183755045,
+      "learning_rate": 5.328398051675423e-06,
+      "loss": 0.5083,
+      "step": 8573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.100667711823135,
+      "learning_rate": 5.32617656401107e-06,
+      "loss": 0.454,
+      "step": 8574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0903712264317869,
+      "learning_rate": 5.323955371429299e-06,
+      "loss": 0.4832,
+      "step": 8575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1942476004420102,
+      "learning_rate": 5.32173447407035e-06,
+      "loss": 0.5588,
+      "step": 8576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.232772704667899,
+      "learning_rate": 5.319513872074442e-06,
+      "loss": 0.5315,
+      "step": 8577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1489232910751808,
+      "learning_rate": 5.31729356558177e-06,
+      "loss": 0.4915,
+      "step": 8578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.219642165893346,
+      "learning_rate": 5.31507355473252e-06,
+      "loss": 0.534,
+      "step": 8579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1251108856655108,
+      "learning_rate": 5.312853839666843e-06,
+      "loss": 0.4948,
+      "step": 8580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.256947047873108,
+      "learning_rate": 5.31063442052489e-06,
+      "loss": 0.5913,
+      "step": 8581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2139167806203217,
+      "learning_rate": 5.308415297446774e-06,
+      "loss": 0.5521,
+      "step": 8582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0642537061444397,
+      "learning_rate": 5.306196470572606e-06,
+      "loss": 0.4871,
+      "step": 8583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1650686332230558,
+      "learning_rate": 5.303977940042477e-06,
+      "loss": 0.5077,
+      "step": 8584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2050793120689047,
+      "learning_rate": 5.301759705996446e-06,
+      "loss": 0.5008,
+      "step": 8585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1789422112523302,
+      "learning_rate": 5.299541768574563e-06,
+      "loss": 0.5113,
+      "step": 8586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0785134762774005,
+      "learning_rate": 5.297324127916858e-06,
+      "loss": 0.5204,
+      "step": 8587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1194235026438761,
+      "learning_rate": 5.2951067841633465e-06,
+      "loss": 0.4781,
+      "step": 8588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1258937146823271,
+      "learning_rate": 5.292889737454019e-06,
+      "loss": 0.497,
+      "step": 8589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1190385755310723,
+      "learning_rate": 5.29067298792885e-06,
+      "loss": 0.5278,
+      "step": 8590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2700620979235986,
+      "learning_rate": 5.288456535727786e-06,
+      "loss": 0.4954,
+      "step": 8591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1660641351829342,
+      "learning_rate": 5.286240380990772e-06,
+      "loss": 0.5957,
+      "step": 8592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1979560790281831,
+      "learning_rate": 5.284024523857721e-06,
+      "loss": 0.5393,
+      "step": 8593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2155599660074705,
+      "learning_rate": 5.281808964468534e-06,
+      "loss": 0.5384,
+      "step": 8594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1032186733166185,
+      "learning_rate": 5.2795937029630905e-06,
+      "loss": 0.5244,
+      "step": 8595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1402868004639521,
+      "learning_rate": 5.277378739481249e-06,
+      "loss": 0.4849,
+      "step": 8596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1043772886432794,
+      "learning_rate": 5.275164074162854e-06,
+      "loss": 0.4652,
+      "step": 8597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1901302320879987,
+      "learning_rate": 5.2729497071477276e-06,
+      "loss": 0.4694,
+      "step": 8598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1750360706555412,
+      "learning_rate": 5.270735638575677e-06,
+      "loss": 0.5124,
+      "step": 8599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1766057972837987,
+      "learning_rate": 5.268521868586487e-06,
+      "loss": 0.5406,
+      "step": 8600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2372698579506118,
+      "learning_rate": 5.266308397319918e-06,
+      "loss": 0.5525,
+      "step": 8601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.219008198310753,
+      "learning_rate": 5.264095224915722e-06,
+      "loss": 0.5339,
+      "step": 8602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2359080391903958,
+      "learning_rate": 5.261882351513629e-06,
+      "loss": 0.4405,
+      "step": 8603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2587329032214187,
+      "learning_rate": 5.259669777253349e-06,
+      "loss": 0.5463,
+      "step": 8604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1094577382356248,
+      "learning_rate": 5.257457502274577e-06,
+      "loss": 0.4718,
+      "step": 8605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1723870239649903,
+      "learning_rate": 5.255245526716976e-06,
+      "loss": 0.5151,
+      "step": 8606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1580931947461246,
+      "learning_rate": 5.253033850720206e-06,
+      "loss": 0.5603,
+      "step": 8607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2049124184908713,
+      "learning_rate": 5.2508224744239e-06,
+      "loss": 0.5852,
+      "step": 8608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.172163253936146,
+      "learning_rate": 5.2486113979676765e-06,
+      "loss": 0.5342,
+      "step": 8609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.23633598337192,
+      "learning_rate": 5.24640062149113e-06,
+      "loss": 0.5674,
+      "step": 8610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.15119312837603,
+      "learning_rate": 5.244190145133834e-06,
+      "loss": 0.4746,
+      "step": 8611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1033367180748352,
+      "learning_rate": 5.241979969035351e-06,
+      "loss": 0.495,
+      "step": 8612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1158579671266355,
+      "learning_rate": 5.239770093335219e-06,
+      "loss": 0.5072,
+      "step": 8613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1623490799593086,
+      "learning_rate": 5.237560518172963e-06,
+      "loss": 0.5323,
+      "step": 8614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1601295339674342,
+      "learning_rate": 5.235351243688085e-06,
+      "loss": 0.5236,
+      "step": 8615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1915982951488509,
+      "learning_rate": 5.233142270020062e-06,
+      "loss": 0.5056,
+      "step": 8616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1061701621864002,
+      "learning_rate": 5.230933597308361e-06,
+      "loss": 0.4866,
+      "step": 8617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1430898710414628,
+      "learning_rate": 5.228725225692426e-06,
+      "loss": 0.46,
+      "step": 8618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0734983490823276,
+      "learning_rate": 5.22651715531169e-06,
+      "loss": 0.4287,
+      "step": 8619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2872246327407593,
+      "learning_rate": 5.224309386305553e-06,
+      "loss": 0.5998,
+      "step": 8620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.129060992026886,
+      "learning_rate": 5.222101918813399e-06,
+      "loss": 0.5322,
+      "step": 8621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2066310478366338,
+      "learning_rate": 5.219894752974602e-06,
+      "loss": 0.5578,
+      "step": 8622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1879721756865282,
+      "learning_rate": 5.217687888928512e-06,
+      "loss": 0.5333,
+      "step": 8623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.119153993004121,
+      "learning_rate": 5.215481326814459e-06,
+      "loss": 0.5662,
+      "step": 8624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.029380433022499,
+      "learning_rate": 5.213275066771759e-06,
+      "loss": 0.4834,
+      "step": 8625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.138691295111383,
+      "learning_rate": 5.211069108939695e-06,
+      "loss": 0.5219,
+      "step": 8626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1652682927023097,
+      "learning_rate": 5.2088634534575465e-06,
+      "loss": 0.5716,
+      "step": 8627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.18220100671412,
+      "learning_rate": 5.206658100464568e-06,
+      "loss": 0.5451,
+      "step": 8628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.170097223301615,
+      "learning_rate": 5.204453050099997e-06,
+      "loss": 0.5406,
+      "step": 8629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.113602280758709,
+      "learning_rate": 5.202248302503047e-06,
+      "loss": 0.4961,
+      "step": 8630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3095343099107075,
+      "learning_rate": 5.200043857812911e-06,
+      "loss": 0.5421,
+      "step": 8631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2304595946925025,
+      "learning_rate": 5.19783971616877e-06,
+      "loss": 0.5503,
+      "step": 8632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.163510902337313,
+      "learning_rate": 5.195635877709783e-06,
+      "loss": 0.4635,
+      "step": 8633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0673613690092136,
+      "learning_rate": 5.193432342575093e-06,
+      "loss": 0.4947,
+      "step": 8634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0948594597110899,
+      "learning_rate": 5.191229110903819e-06,
+      "loss": 0.5048,
+      "step": 8635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1360119591909523,
+      "learning_rate": 5.189026182835059e-06,
+      "loss": 0.5391,
+      "step": 8636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.129882073958542,
+      "learning_rate": 5.186823558507897e-06,
+      "loss": 0.5073,
+      "step": 8637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1633678642292653,
+      "learning_rate": 5.184621238061397e-06,
+      "loss": 0.5447,
+      "step": 8638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1809202410084578,
+      "learning_rate": 5.182419221634605e-06,
+      "loss": 0.5435,
+      "step": 8639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.160708671598167,
+      "learning_rate": 5.180217509366544e-06,
+      "loss": 0.5088,
+      "step": 8640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1131224401874014,
+      "learning_rate": 5.178016101396215e-06,
+      "loss": 0.5037,
+      "step": 8641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2057839773200891,
+      "learning_rate": 5.175814997862606e-06,
+      "loss": 0.5706,
+      "step": 8642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1948588888097982,
+      "learning_rate": 5.173614198904686e-06,
+      "loss": 0.5504,
+      "step": 8643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1293594876075312,
+      "learning_rate": 5.171413704661403e-06,
+      "loss": 0.5333,
+      "step": 8644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2648609822296937,
+      "learning_rate": 5.169213515271686e-06,
+      "loss": 0.5791,
+      "step": 8645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.088862315624611,
+      "learning_rate": 5.167013630874447e-06,
+      "loss": 0.4896,
+      "step": 8646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2043051938076799,
+      "learning_rate": 5.164814051608567e-06,
+      "loss": 0.5396,
+      "step": 8647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2169972556309723,
+      "learning_rate": 5.162614777612924e-06,
+      "loss": 0.5411,
+      "step": 8648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1343408731839044,
+      "learning_rate": 5.16041580902637e-06,
+      "loss": 0.547,
+      "step": 8649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1817151258958554,
+      "learning_rate": 5.158217145987732e-06,
+      "loss": 0.5318,
+      "step": 8650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.097955574682945,
+      "learning_rate": 5.156018788635831e-06,
+      "loss": 0.5265,
+      "step": 8651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.198500278226194,
+      "learning_rate": 5.153820737109449e-06,
+      "loss": 0.519,
+      "step": 8652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.212108120070904,
+      "learning_rate": 5.151622991547368e-06,
+      "loss": 0.5715,
+      "step": 8653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1006655456942376,
+      "learning_rate": 5.149425552088342e-06,
+      "loss": 0.4883,
+      "step": 8654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0808006477462726,
+      "learning_rate": 5.1472284188711065e-06,
+      "loss": 0.497,
+      "step": 8655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.151524553230149,
+      "learning_rate": 5.145031592034382e-06,
+      "loss": 0.5578,
+      "step": 8656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1521271582897061,
+      "learning_rate": 5.1428350717168575e-06,
+      "loss": 0.4992,
+      "step": 8657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1410046820816142,
+      "learning_rate": 5.140638858057214e-06,
+      "loss": 0.5163,
+      "step": 8658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2454646803703624,
+      "learning_rate": 5.138442951194115e-06,
+      "loss": 0.5194,
+      "step": 8659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1634886178286548,
+      "learning_rate": 5.136247351266191e-06,
+      "loss": 0.4653,
+      "step": 8660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1077306138780623,
+      "learning_rate": 5.134052058412069e-06,
+      "loss": 0.4784,
+      "step": 8661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.219274505822627,
+      "learning_rate": 5.131857072770341e-06,
+      "loss": 0.5251,
+      "step": 8662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1778286568780172,
+      "learning_rate": 5.129662394479593e-06,
+      "loss": 0.5141,
+      "step": 8663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2318246304120248,
+      "learning_rate": 5.1274680236783855e-06,
+      "loss": 0.5268,
+      "step": 8664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0200855604451335,
+      "learning_rate": 5.125273960505261e-06,
+      "loss": 0.4815,
+      "step": 8665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1186225584103466,
+      "learning_rate": 5.123080205098745e-06,
+      "loss": 0.5091,
+      "step": 8666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.196980229226421,
+      "learning_rate": 5.120886757597334e-06,
+      "loss": 0.5641,
+      "step": 8667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0676452928246292,
+      "learning_rate": 5.118693618139514e-06,
+      "loss": 0.4918,
+      "step": 8668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1282094633231983,
+      "learning_rate": 5.116500786863755e-06,
+      "loss": 0.5331,
+      "step": 8669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1163498198847663,
+      "learning_rate": 5.114308263908493e-06,
+      "loss": 0.5015,
+      "step": 8670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2569282219239768,
+      "learning_rate": 5.11211604941216e-06,
+      "loss": 0.5146,
+      "step": 8671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.024830459901714,
+      "learning_rate": 5.109924143513156e-06,
+      "loss": 0.4596,
+      "step": 8672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1936340515387955,
+      "learning_rate": 5.107732546349871e-06,
+      "loss": 0.5151,
+      "step": 8673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.217341415996466,
+      "learning_rate": 5.105541258060669e-06,
+      "loss": 0.5312,
+      "step": 8674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1673112121957276,
+      "learning_rate": 5.103350278783901e-06,
+      "loss": 0.5564,
+      "step": 8675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2313955056436072,
+      "learning_rate": 5.101159608657896e-06,
+      "loss": 0.4974,
+      "step": 8676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.094801206876865,
+      "learning_rate": 5.098969247820958e-06,
+      "loss": 0.5384,
+      "step": 8677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2621264669948749,
+      "learning_rate": 5.096779196411375e-06,
+      "loss": 0.5435,
+      "step": 8678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1854095630846337,
+      "learning_rate": 5.094589454567423e-06,
+      "loss": 0.4999,
+      "step": 8679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2592366845443836,
+      "learning_rate": 5.092400022427344e-06,
+      "loss": 0.5658,
+      "step": 8680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1612638073973949,
+      "learning_rate": 5.090210900129375e-06,
+      "loss": 0.4336,
+      "step": 8681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1419678713025245,
+      "learning_rate": 5.088022087811719e-06,
+      "loss": 0.5373,
+      "step": 8682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1215963267244438,
+      "learning_rate": 5.0858335856125715e-06,
+      "loss": 0.4958,
+      "step": 8683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0964654728558387,
+      "learning_rate": 5.083645393670103e-06,
+      "loss": 0.4404,
+      "step": 8684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.107117626236695,
+      "learning_rate": 5.081457512122466e-06,
+      "loss": 0.4801,
+      "step": 8685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1830450114170197,
+      "learning_rate": 5.0792699411077976e-06,
+      "loss": 0.4896,
+      "step": 8686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2074752040552232,
+      "learning_rate": 5.077082680764201e-06,
+      "loss": 0.5487,
+      "step": 8687
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2304433668636838,
+      "learning_rate": 5.074895731229772e-06,
+      "loss": 0.5596,
+      "step": 8688
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1889513582878708,
+      "learning_rate": 5.0727090926425915e-06,
+      "loss": 0.5316,
+      "step": 8689
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.333391540965462,
+      "learning_rate": 5.070522765140703e-06,
+      "loss": 0.6063,
+      "step": 8690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2302187711104398,
+      "learning_rate": 5.068336748862148e-06,
+      "loss": 0.5517,
+      "step": 8691
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.15457893449672,
+      "learning_rate": 5.066151043944936e-06,
+      "loss": 0.5197,
+      "step": 8692
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3483178767499935,
+      "learning_rate": 5.063965650527063e-06,
+      "loss": 0.4857,
+      "step": 8693
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1334459835520658,
+      "learning_rate": 5.0617805687465064e-06,
+      "loss": 0.5382,
+      "step": 8694
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3725335801816276,
+      "learning_rate": 5.059595798741218e-06,
+      "loss": 0.5671,
+      "step": 8695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2457734177555044,
+      "learning_rate": 5.0574113406491365e-06,
+      "loss": 0.5113,
+      "step": 8696
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.216837531768351,
+      "learning_rate": 5.055227194608183e-06,
+      "loss": 0.5538,
+      "step": 8697
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1249549644780816,
+      "learning_rate": 5.053043360756247e-06,
+      "loss": 0.5618,
+      "step": 8698
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2482107708552286,
+      "learning_rate": 5.050859839231203e-06,
+      "loss": 0.4982,
+      "step": 8699
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2062114412934317,
+      "learning_rate": 5.0486766301709115e-06,
+      "loss": 0.5312,
+      "step": 8700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1889998851837025,
+      "learning_rate": 5.046493733713209e-06,
+      "loss": 0.5058,
+      "step": 8701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2844974951328014,
+      "learning_rate": 5.04431114999592e-06,
+      "loss": 0.5222,
+      "step": 8702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2528252621288747,
+      "learning_rate": 5.0421288791568305e-06,
+      "loss": 0.5214,
+      "step": 8703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1496130188531966,
+      "learning_rate": 5.0399469213337234e-06,
+      "loss": 0.5005,
+      "step": 8704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2741424855810366,
+      "learning_rate": 5.037765276664359e-06,
+      "loss": 0.5238,
+      "step": 8705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1740731032874818,
+      "learning_rate": 5.035583945286474e-06,
+      "loss": 0.5199,
+      "step": 8706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2949348552622904,
+      "learning_rate": 5.033402927337792e-06,
+      "loss": 0.5813,
+      "step": 8707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2079318738920737,
+      "learning_rate": 5.031222222956007e-06,
+      "loss": 0.5605,
+      "step": 8708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.186743746425818,
+      "learning_rate": 5.029041832278794e-06,
+      "loss": 0.5521,
+      "step": 8709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1617586520229715,
+      "learning_rate": 5.026861755443817e-06,
+      "loss": 0.4885,
+      "step": 8710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.225214028637384,
+      "learning_rate": 5.024681992588717e-06,
+      "loss": 0.535,
+      "step": 8711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0467486518794695,
+      "learning_rate": 5.022502543851116e-06,
+      "loss": 0.49,
+      "step": 8712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1716284937674728,
+      "learning_rate": 5.020323409368604e-06,
+      "loss": 0.5247,
+      "step": 8713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.105648390121989,
+      "learning_rate": 5.018144589278768e-06,
+      "loss": 0.4994,
+      "step": 8714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0612254070941622,
+      "learning_rate": 5.015966083719166e-06,
+      "loss": 0.4692,
+      "step": 8715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2910441108365716,
+      "learning_rate": 5.013787892827341e-06,
+      "loss": 0.5977,
+      "step": 8716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1590311779604443,
+      "learning_rate": 5.011610016740815e-06,
+      "loss": 0.5083,
+      "step": 8717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2803542564832306,
+      "learning_rate": 5.009432455597085e-06,
+      "loss": 0.5506,
+      "step": 8718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.31318151854397,
+      "learning_rate": 5.007255209533629e-06,
+      "loss": 0.5438,
+      "step": 8719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.226707692876925,
+      "learning_rate": 5.00507827868791e-06,
+      "loss": 0.5622,
+      "step": 8720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1978867181544943,
+      "learning_rate": 5.00290166319737e-06,
+      "loss": 0.5242,
+      "step": 8721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1723747205196022,
+      "learning_rate": 5.000725363199433e-06,
+      "loss": 0.4761,
+      "step": 8722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1578956955068445,
+      "learning_rate": 4.998549378831494e-06,
+      "loss": 0.556,
+      "step": 8723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1441949864656908,
+      "learning_rate": 4.996373710230937e-06,
+      "loss": 0.4635,
+      "step": 8724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.087636209047633,
+      "learning_rate": 4.994198357535122e-06,
+      "loss": 0.463,
+      "step": 8725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1115647317189916,
+      "learning_rate": 4.992023320881391e-06,
+      "loss": 0.5276,
+      "step": 8726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1090953098236815,
+      "learning_rate": 4.989848600407069e-06,
+      "loss": 0.5932,
+      "step": 8727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2186306626018797,
+      "learning_rate": 4.987674196249454e-06,
+      "loss": 0.5357,
+      "step": 8728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.113205649502695,
+      "learning_rate": 4.9855001085458246e-06,
+      "loss": 0.5084,
+      "step": 8729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.35781143416702,
+      "learning_rate": 4.983326337433444e-06,
+      "loss": 0.5489,
+      "step": 8730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1932374482996237,
+      "learning_rate": 4.981152883049555e-06,
+      "loss": 0.557,
+      "step": 8731
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2348166532752183,
+      "learning_rate": 4.978979745531382e-06,
+      "loss": 0.5066,
+      "step": 8732
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.163448453406435,
+      "learning_rate": 4.976806925016117e-06,
+      "loss": 0.5448,
+      "step": 8733
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.245600827102656,
+      "learning_rate": 4.974634421640949e-06,
+      "loss": 0.5817,
+      "step": 8734
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2253249905063204,
+      "learning_rate": 4.972462235543038e-06,
+      "loss": 0.5275,
+      "step": 8735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2104947665171129,
+      "learning_rate": 4.970290366859523e-06,
+      "loss": 0.5017,
+      "step": 8736
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1282664667335658,
+      "learning_rate": 4.968118815727532e-06,
+      "loss": 0.521,
+      "step": 8737
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1283011217088237,
+      "learning_rate": 4.965947582284161e-06,
+      "loss": 0.4936,
+      "step": 8738
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1676207569413988,
+      "learning_rate": 4.963776666666487e-06,
+      "loss": 0.5043,
+      "step": 8739
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.193596249810098,
+      "learning_rate": 4.961606069011576e-06,
+      "loss": 0.4662,
+      "step": 8740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.233221166682347,
+      "learning_rate": 4.959435789456468e-06,
+      "loss": 0.5592,
+      "step": 8741
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.144769689310219,
+      "learning_rate": 4.957265828138189e-06,
+      "loss": 0.5146,
+      "step": 8742
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1987768595434038,
+      "learning_rate": 4.955096185193732e-06,
+      "loss": 0.5294,
+      "step": 8743
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2877682286081171,
+      "learning_rate": 4.95292686076008e-06,
+      "loss": 0.5436,
+      "step": 8744
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2587809654904685,
+      "learning_rate": 4.950757854974195e-06,
+      "loss": 0.5423,
+      "step": 8745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2094043501414709,
+      "learning_rate": 4.948589167973018e-06,
+      "loss": 0.5208,
+      "step": 8746
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4005826555053236,
+      "learning_rate": 4.946420799893472e-06,
+      "loss": 0.489,
+      "step": 8747
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0945550408184055,
+      "learning_rate": 4.944252750872455e-06,
+      "loss": 0.5158,
+      "step": 8748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.208639758269382,
+      "learning_rate": 4.94208502104684e-06,
+      "loss": 0.5094,
+      "step": 8749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2441755496502385,
+      "learning_rate": 4.9399176105534954e-06,
+      "loss": 0.5618,
+      "step": 8750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1204868494885198,
+      "learning_rate": 4.937750519529258e-06,
+      "loss": 0.4654,
+      "step": 8751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0849894494980843,
+      "learning_rate": 4.935583748110947e-06,
+      "loss": 0.484,
+      "step": 8752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1989675749625017,
+      "learning_rate": 4.933417296435367e-06,
+      "loss": 0.5035,
+      "step": 8753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1804838208613058,
+      "learning_rate": 4.931251164639289e-06,
+      "loss": 0.4876,
+      "step": 8754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2706302522249588,
+      "learning_rate": 4.929085352859478e-06,
+      "loss": 0.5765,
+      "step": 8755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.110206238858068,
+      "learning_rate": 4.92691986123267e-06,
+      "loss": 0.4961,
+      "step": 8756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1098562459391241,
+      "learning_rate": 4.924754689895589e-06,
+      "loss": 0.4791,
+      "step": 8757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1283116342119364,
+      "learning_rate": 4.922589838984929e-06,
+      "loss": 0.5065,
+      "step": 8758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.198320530799597,
+      "learning_rate": 4.920425308637365e-06,
+      "loss": 0.5839,
+      "step": 8759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2064124437757886,
+      "learning_rate": 4.918261098989557e-06,
+      "loss": 0.5306,
+      "step": 8760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1800092548880279,
+      "learning_rate": 4.9160972101781455e-06,
+      "loss": 0.5327,
+      "step": 8761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1973206848129216,
+      "learning_rate": 4.913933642339747e-06,
+      "loss": 0.522,
+      "step": 8762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1384549335042284,
+      "learning_rate": 4.911770395610961e-06,
+      "loss": 0.4791,
+      "step": 8763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0463926997794826,
+      "learning_rate": 4.909607470128358e-06,
+      "loss": 0.4773,
+      "step": 8764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2602889519082285,
+      "learning_rate": 4.907444866028499e-06,
+      "loss": 0.492,
+      "step": 8765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1354075731852413,
+      "learning_rate": 4.90528258344792e-06,
+      "loss": 0.5507,
+      "step": 8766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1896138200121311,
+      "learning_rate": 4.90312062252314e-06,
+      "loss": 0.5114,
+      "step": 8767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1859678121319162,
+      "learning_rate": 4.900958983390651e-06,
+      "loss": 0.4944,
+      "step": 8768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3874209922771823,
+      "learning_rate": 4.898797666186926e-06,
+      "loss": 0.6106,
+      "step": 8769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1651552946524932,
+      "learning_rate": 4.896636671048424e-06,
+      "loss": 0.514,
+      "step": 8770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2597598053553805,
+      "learning_rate": 4.894475998111578e-06,
+      "loss": 0.4905,
+      "step": 8771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1090669875969148,
+      "learning_rate": 4.892315647512802e-06,
+      "loss": 0.4978,
+      "step": 8772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3318073610327208,
+      "learning_rate": 4.890155619388497e-06,
+      "loss": 0.5835,
+      "step": 8773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1785955364684588,
+      "learning_rate": 4.887995913875025e-06,
+      "loss": 0.5015,
+      "step": 8774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.270471594350697,
+      "learning_rate": 4.8858365311087475e-06,
+      "loss": 0.5193,
+      "step": 8775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1723017614858855,
+      "learning_rate": 4.883677471225995e-06,
+      "loss": 0.5309,
+      "step": 8776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0159863342732784,
+      "learning_rate": 4.881518734363084e-06,
+      "loss": 0.4668,
+      "step": 8777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1852034396569366,
+      "learning_rate": 4.8793603206563034e-06,
+      "loss": 0.5226,
+      "step": 8778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1410620387744448,
+      "learning_rate": 4.8772022302419206e-06,
+      "loss": 0.5536,
+      "step": 8779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0879014732052814,
+      "learning_rate": 4.875044463256192e-06,
+      "loss": 0.5078,
+      "step": 8780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1385734083678445,
+      "learning_rate": 4.872887019835347e-06,
+      "loss": 0.5632,
+      "step": 8781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1599100795785098,
+      "learning_rate": 4.870729900115597e-06,
+      "loss": 0.5211,
+      "step": 8782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1990562602246906,
+      "learning_rate": 4.868573104233137e-06,
+      "loss": 0.4641,
+      "step": 8783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1820757608548764,
+      "learning_rate": 4.8664166323241255e-06,
+      "loss": 0.5194,
+      "step": 8784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.151315676610267,
+      "learning_rate": 4.864260484524719e-06,
+      "loss": 0.5376,
+      "step": 8785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2034011127653805,
+      "learning_rate": 4.862104660971045e-06,
+      "loss": 0.4728,
+      "step": 8786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.233055375111033,
+      "learning_rate": 4.859949161799216e-06,
+      "loss": 0.5638,
+      "step": 8787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1503347179253807,
+      "learning_rate": 4.857793987145315e-06,
+      "loss": 0.5237,
+      "step": 8788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.191761251621735,
+      "learning_rate": 4.855639137145406e-06,
+      "loss": 0.5091,
+      "step": 8789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1556485648096828,
+      "learning_rate": 4.853484611935541e-06,
+      "loss": 0.5012,
+      "step": 8790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2088060386157224,
+      "learning_rate": 4.8513304116517435e-06,
+      "loss": 0.5373,
+      "step": 8791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1651812815934823,
+      "learning_rate": 4.8491765364300205e-06,
+      "loss": 0.5175,
+      "step": 8792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3091126736814878,
+      "learning_rate": 4.847022986406362e-06,
+      "loss": 0.5356,
+      "step": 8793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1208746293885075,
+      "learning_rate": 4.844869761716725e-06,
+      "loss": 0.4709,
+      "step": 8794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.241821523051576,
+      "learning_rate": 4.842716862497056e-06,
+      "loss": 0.51,
+      "step": 8795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1208106025650384,
+      "learning_rate": 4.840564288883284e-06,
+      "loss": 0.4849,
+      "step": 8796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3080502876128315,
+      "learning_rate": 4.838412041011304e-06,
+      "loss": 0.571,
+      "step": 8797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.146630402969278,
+      "learning_rate": 4.836260119017005e-06,
+      "loss": 0.5836,
+      "step": 8798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2412763888962388,
+      "learning_rate": 4.8341085230362425e-06,
+      "loss": 0.4665,
+      "step": 8799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1507169976202056,
+      "learning_rate": 4.831957253204862e-06,
+      "loss": 0.5157,
+      "step": 8800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.011558549488498,
+      "learning_rate": 4.829806309658683e-06,
+      "loss": 0.4807,
+      "step": 8801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1604659569417641,
+      "learning_rate": 4.827655692533506e-06,
+      "loss": 0.4749,
+      "step": 8802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.110122321494262,
+      "learning_rate": 4.825505401965116e-06,
+      "loss": 0.4615,
+      "step": 8803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1882902326877618,
+      "learning_rate": 4.823355438089262e-06,
+      "loss": 0.4703,
+      "step": 8804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2340727810348104,
+      "learning_rate": 4.821205801041688e-06,
+      "loss": 0.5237,
+      "step": 8805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2388076866567943,
+      "learning_rate": 4.819056490958115e-06,
+      "loss": 0.514,
+      "step": 8806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1694750816296748,
+      "learning_rate": 4.816907507974231e-06,
+      "loss": 0.5062,
+      "step": 8807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3178097179547208,
+      "learning_rate": 4.814758852225717e-06,
+      "loss": 0.5395,
+      "step": 8808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3070496516234882,
+      "learning_rate": 4.8126105238482345e-06,
+      "loss": 0.5812,
+      "step": 8809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2633119804688122,
+      "learning_rate": 4.810462522977408e-06,
+      "loss": 0.5497,
+      "step": 8810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1145987762141907,
+      "learning_rate": 4.808314849748858e-06,
+      "loss": 0.5404,
+      "step": 8811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2025891386370484,
+      "learning_rate": 4.806167504298175e-06,
+      "loss": 0.5238,
+      "step": 8812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.16903076879311,
+      "learning_rate": 4.8040204867609355e-06,
+      "loss": 0.499,
+      "step": 8813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3988739270357924,
+      "learning_rate": 4.801873797272694e-06,
+      "loss": 0.61,
+      "step": 8814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1930018512203053,
+      "learning_rate": 4.799727435968975e-06,
+      "loss": 0.553,
+      "step": 8815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.133625501717757,
+      "learning_rate": 4.797581402985296e-06,
+      "loss": 0.4546,
+      "step": 8816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2365295824552878,
+      "learning_rate": 4.795435698457141e-06,
+      "loss": 0.5232,
+      "step": 8817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.20709993728702,
+      "learning_rate": 4.793290322519981e-06,
+      "loss": 0.5143,
+      "step": 8818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.108837857297423,
+      "learning_rate": 4.791145275309271e-06,
+      "loss": 0.5246,
+      "step": 8819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1630115757511255,
+      "learning_rate": 4.7890005569604305e-06,
+      "loss": 0.4973,
+      "step": 8820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2096021118616156,
+      "learning_rate": 4.786856167608869e-06,
+      "loss": 0.4928,
+      "step": 8821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.292562830152313,
+      "learning_rate": 4.784712107389975e-06,
+      "loss": 0.523,
+      "step": 8822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1632057467745005,
+      "learning_rate": 4.7825683764391114e-06,
+      "loss": 0.526,
+      "step": 8823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.175446877824334,
+      "learning_rate": 4.780424974891629e-06,
+      "loss": 0.5266,
+      "step": 8824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1399257031707766,
+      "learning_rate": 4.778281902882844e-06,
+      "loss": 0.4544,
+      "step": 8825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3028203429720102,
+      "learning_rate": 4.776139160548068e-06,
+      "loss": 0.5362,
+      "step": 8826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3395818337550256,
+      "learning_rate": 4.773996748022573e-06,
+      "loss": 0.5655,
+      "step": 8827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2426064222723974,
+      "learning_rate": 4.771854665441626e-06,
+      "loss": 0.5484,
+      "step": 8828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1789879651122728,
+      "learning_rate": 4.769712912940472e-06,
+      "loss": 0.5561,
+      "step": 8829
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1415945139358161,
+      "learning_rate": 4.767571490654322e-06,
+      "loss": 0.4989,
+      "step": 8830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2775170911478029,
+      "learning_rate": 4.76543039871838e-06,
+      "loss": 0.542,
+      "step": 8831
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1832591172348037,
+      "learning_rate": 4.7632896372678235e-06,
+      "loss": 0.5082,
+      "step": 8832
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.179159741106128,
+      "learning_rate": 4.761149206437811e-06,
+      "loss": 0.5019,
+      "step": 8833
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.17375779536282,
+      "learning_rate": 4.759009106363482e-06,
+      "loss": 0.5362,
+      "step": 8834
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1856792947888322,
+      "learning_rate": 4.756869337179942e-06,
+      "loss": 0.5244,
+      "step": 8835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.247044550331615,
+      "learning_rate": 4.754729899022298e-06,
+      "loss": 0.5349,
+      "step": 8836
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1721126061201548,
+      "learning_rate": 4.752590792025612e-06,
+      "loss": 0.4768,
+      "step": 8837
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2372419646999036,
+      "learning_rate": 4.750452016324943e-06,
+      "loss": 0.5551,
+      "step": 8838
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.201077837480701,
+      "learning_rate": 4.748313572055326e-06,
+      "loss": 0.5171,
+      "step": 8839
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0790756015330216,
+      "learning_rate": 4.746175459351765e-06,
+      "loss": 0.4984,
+      "step": 8840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1799470224449369,
+      "learning_rate": 4.744037678349254e-06,
+      "loss": 0.5413,
+      "step": 8841
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1348939911662688,
+      "learning_rate": 4.74190022918276e-06,
+      "loss": 0.514,
+      "step": 8842
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.059176351551484,
+      "learning_rate": 4.7397631119872335e-06,
+      "loss": 0.5196,
+      "step": 8843
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1550405207870562,
+      "learning_rate": 4.737626326897604e-06,
+      "loss": 0.516,
+      "step": 8844
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1138553678289136,
+      "learning_rate": 4.735489874048772e-06,
+      "loss": 0.5321,
+      "step": 8845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1132218729882424,
+      "learning_rate": 4.733353753575629e-06,
+      "loss": 0.4097,
+      "step": 8846
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4144234381054221,
+      "learning_rate": 4.731217965613033e-06,
+      "loss": 0.5674,
+      "step": 8847
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1599987706638268,
+      "learning_rate": 4.729082510295829e-06,
+      "loss": 0.5178,
+      "step": 8848
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2147427728520006,
+      "learning_rate": 4.726947387758845e-06,
+      "loss": 0.4921,
+      "step": 8849
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2836594604835394,
+      "learning_rate": 4.724812598136873e-06,
+      "loss": 0.5406,
+      "step": 8850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.155998821297543,
+      "learning_rate": 4.722678141564698e-06,
+      "loss": 0.521,
+      "step": 8851
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0403337661068388,
+      "learning_rate": 4.720544018177081e-06,
+      "loss": 0.4557,
+      "step": 8852
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.257351998327825,
+      "learning_rate": 4.718410228108757e-06,
+      "loss": 0.5171,
+      "step": 8853
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2262320589703215,
+      "learning_rate": 4.716276771494449e-06,
+      "loss": 0.4803,
+      "step": 8854
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2557203533296322,
+      "learning_rate": 4.714143648468845e-06,
+      "loss": 0.5221,
+      "step": 8855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1901707481051114,
+      "learning_rate": 4.712010859166628e-06,
+      "loss": 0.5721,
+      "step": 8856
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.277576810240922,
+      "learning_rate": 4.709878403722445e-06,
+      "loss": 0.5115,
+      "step": 8857
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1938510013213617,
+      "learning_rate": 4.70774628227093e-06,
+      "loss": 0.5465,
+      "step": 8858
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.17692839466426,
+      "learning_rate": 4.7056144949467005e-06,
+      "loss": 0.5193,
+      "step": 8859
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1323715404049608,
+      "learning_rate": 4.70348304188434e-06,
+      "loss": 0.5282,
+      "step": 8860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1541353425734135,
+      "learning_rate": 4.7013519232184225e-06,
+      "loss": 0.5003,
+      "step": 8861
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3676767400311258,
+      "learning_rate": 4.699221139083494e-06,
+      "loss": 0.5499,
+      "step": 8862
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2077386258568266,
+      "learning_rate": 4.697090689614084e-06,
+      "loss": 0.5235,
+      "step": 8863
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2054689042690176,
+      "learning_rate": 4.6949605749446995e-06,
+      "loss": 0.5145,
+      "step": 8864
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2237714367743355,
+      "learning_rate": 4.692830795209826e-06,
+      "loss": 0.5602,
+      "step": 8865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1963040596158168,
+      "learning_rate": 4.690701350543927e-06,
+      "loss": 0.5039,
+      "step": 8866
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.225439298451685,
+      "learning_rate": 4.6885722410814396e-06,
+      "loss": 0.5547,
+      "step": 8867
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2616693823636926,
+      "learning_rate": 4.68644346695679e-06,
+      "loss": 0.5171,
+      "step": 8868
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3255002283119859,
+      "learning_rate": 4.684315028304379e-06,
+      "loss": 0.5759,
+      "step": 8869
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2304056787220903,
+      "learning_rate": 4.68218692525859e-06,
+      "loss": 0.5558,
+      "step": 8870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2182677855560362,
+      "learning_rate": 4.680059157953772e-06,
+      "loss": 0.5184,
+      "step": 8871
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.218183875033662,
+      "learning_rate": 4.6779317265242675e-06,
+      "loss": 0.522,
+      "step": 8872
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.226259230521611,
+      "learning_rate": 4.6758046311043915e-06,
+      "loss": 0.574,
+      "step": 8873
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2420230972611777,
+      "learning_rate": 4.673677871828437e-06,
+      "loss": 0.5738,
+      "step": 8874
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2252305689658154,
+      "learning_rate": 4.671551448830684e-06,
+      "loss": 0.5285,
+      "step": 8875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2394194561424576,
+      "learning_rate": 4.669425362245378e-06,
+      "loss": 0.5243,
+      "step": 8876
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2249773801933952,
+      "learning_rate": 4.667299612206747e-06,
+      "loss": 0.5395,
+      "step": 8877
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2668172142426282,
+      "learning_rate": 4.665174198849006e-06,
+      "loss": 0.541,
+      "step": 8878
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2793023728114212,
+      "learning_rate": 4.6630491223063415e-06,
+      "loss": 0.5331,
+      "step": 8879
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1886959077259043,
+      "learning_rate": 4.660924382712923e-06,
+      "loss": 0.5406,
+      "step": 8880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.137707467075631,
+      "learning_rate": 4.658799980202893e-06,
+      "loss": 0.4997,
+      "step": 8881
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2855292709383201,
+      "learning_rate": 4.656675914910376e-06,
+      "loss": 0.5292,
+      "step": 8882
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2334257411275467,
+      "learning_rate": 4.654552186969477e-06,
+      "loss": 0.5588,
+      "step": 8883
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1992880338355738,
+      "learning_rate": 4.6524287965142765e-06,
+      "loss": 0.5029,
+      "step": 8884
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2504801781573172,
+      "learning_rate": 4.6503057436788405e-06,
+      "loss": 0.5656,
+      "step": 8885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2238899320062235,
+      "learning_rate": 4.648183028597203e-06,
+      "loss": 0.513,
+      "step": 8886
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2239236813232504,
+      "learning_rate": 4.646060651403379e-06,
+      "loss": 0.4781,
+      "step": 8887
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2229128924717194,
+      "learning_rate": 4.643938612231369e-06,
+      "loss": 0.5468,
+      "step": 8888
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1652624614730998,
+      "learning_rate": 4.641816911215148e-06,
+      "loss": 0.489,
+      "step": 8889
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1466826441302262,
+      "learning_rate": 4.639695548488673e-06,
+      "loss": 0.482,
+      "step": 8890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.181788209888564,
+      "learning_rate": 4.63757452418587e-06,
+      "loss": 0.5463,
+      "step": 8891
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2867847548453584,
+      "learning_rate": 4.635453838440654e-06,
+      "loss": 0.5273,
+      "step": 8892
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.089022255021199,
+      "learning_rate": 4.633333491386915e-06,
+      "loss": 0.4706,
+      "step": 8893
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1810854782095423,
+      "learning_rate": 4.631213483158525e-06,
+      "loss": 0.475,
+      "step": 8894
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1169802066582672,
+      "learning_rate": 4.6290938138893225e-06,
+      "loss": 0.5131,
+      "step": 8895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2482594292634268,
+      "learning_rate": 4.626974483713142e-06,
+      "loss": 0.5399,
+      "step": 8896
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1119416861132356,
+      "learning_rate": 4.62485549276378e-06,
+      "loss": 0.5233,
+      "step": 8897
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2294986352001556,
+      "learning_rate": 4.622736841175023e-06,
+      "loss": 0.5361,
+      "step": 8898
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.146167405874918,
+      "learning_rate": 4.620618529080632e-06,
+      "loss": 0.5348,
+      "step": 8899
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1541456197662716,
+      "learning_rate": 4.61850055661435e-06,
+      "loss": 0.4828,
+      "step": 8900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.208325528525216,
+      "learning_rate": 4.61638292390989e-06,
+      "loss": 0.5133,
+      "step": 8901
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2489954249601358,
+      "learning_rate": 4.614265631100952e-06,
+      "loss": 0.5533,
+      "step": 8902
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.311399679814041,
+      "learning_rate": 4.612148678321211e-06,
+      "loss": 0.5022,
+      "step": 8903
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1329007344577136,
+      "learning_rate": 4.610032065704325e-06,
+      "loss": 0.5266,
+      "step": 8904
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.213177092749023,
+      "learning_rate": 4.607915793383921e-06,
+      "loss": 0.5389,
+      "step": 8905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2040836923824745,
+      "learning_rate": 4.605799861493615e-06,
+      "loss": 0.4958,
+      "step": 8906
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0646999807832698,
+      "learning_rate": 4.60368427016699e-06,
+      "loss": 0.5065,
+      "step": 8907
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1537965571308,
+      "learning_rate": 4.601569019537619e-06,
+      "loss": 0.5029,
+      "step": 8908
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.199606916499002,
+      "learning_rate": 4.599454109739047e-06,
+      "loss": 0.4962,
+      "step": 8909
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.193114659867559,
+      "learning_rate": 4.597339540904806e-06,
+      "loss": 0.5113,
+      "step": 8910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2220495991950413,
+      "learning_rate": 4.595225313168389e-06,
+      "loss": 0.5851,
+      "step": 8911
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2444790510113326,
+      "learning_rate": 4.5931114266632826e-06,
+      "loss": 0.5201,
+      "step": 8912
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1684836078941885,
+      "learning_rate": 4.590997881522948e-06,
+      "loss": 0.4916,
+      "step": 8913
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1607906777531398,
+      "learning_rate": 4.588884677880828e-06,
+      "loss": 0.4283,
+      "step": 8914
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1203906682408602,
+      "learning_rate": 4.586771815870332e-06,
+      "loss": 0.5341,
+      "step": 8915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1136149123926693,
+      "learning_rate": 4.5846592956248636e-06,
+      "loss": 0.5302,
+      "step": 8916
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1012851014101719,
+      "learning_rate": 4.582547117277789e-06,
+      "loss": 0.45,
+      "step": 8917
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1800675443132653,
+      "learning_rate": 4.580435280962466e-06,
+      "loss": 0.5559,
+      "step": 8918
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.06715070848172,
+      "learning_rate": 4.578323786812225e-06,
+      "loss": 0.495,
+      "step": 8919
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2273294760045803,
+      "learning_rate": 4.5762126349603755e-06,
+      "loss": 0.5142,
+      "step": 8920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1333739895166621,
+      "learning_rate": 4.574101825540209e-06,
+      "loss": 0.4715,
+      "step": 8921
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.241579062723198,
+      "learning_rate": 4.571991358684984e-06,
+      "loss": 0.5607,
+      "step": 8922
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1906023158518209,
+      "learning_rate": 4.569881234527951e-06,
+      "loss": 0.5334,
+      "step": 8923
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2194152874648232,
+      "learning_rate": 4.5677714532023335e-06,
+      "loss": 0.5325,
+      "step": 8924
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1392613450624556,
+      "learning_rate": 4.565662014841328e-06,
+      "loss": 0.5424,
+      "step": 8925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2520186813258727,
+      "learning_rate": 4.563552919578119e-06,
+      "loss": 0.4893,
+      "step": 8926
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1390535686545693,
+      "learning_rate": 4.561444167545861e-06,
+      "loss": 0.5007,
+      "step": 8927
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1405801633618593,
+      "learning_rate": 4.559335758877691e-06,
+      "loss": 0.5332,
+      "step": 8928
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1148411045698017,
+      "learning_rate": 4.557227693706724e-06,
+      "loss": 0.5014,
+      "step": 8929
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0837685859960804,
+      "learning_rate": 4.555119972166055e-06,
+      "loss": 0.4503,
+      "step": 8930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.194008957901136,
+      "learning_rate": 4.553012594388757e-06,
+      "loss": 0.4709,
+      "step": 8931
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1867817664225282,
+      "learning_rate": 4.550905560507873e-06,
+      "loss": 0.4792,
+      "step": 8932
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.126105083664196,
+      "learning_rate": 4.548798870656434e-06,
+      "loss": 0.4946,
+      "step": 8933
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.207783930392495,
+      "learning_rate": 4.546692524967452e-06,
+      "loss": 0.4909,
+      "step": 8934
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2831898867128917,
+      "learning_rate": 4.544586523573902e-06,
+      "loss": 0.5173,
+      "step": 8935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1978771645487492,
+      "learning_rate": 4.542480866608754e-06,
+      "loss": 0.4947,
+      "step": 8936
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2085342185432035,
+      "learning_rate": 4.540375554204944e-06,
+      "loss": 0.4989,
+      "step": 8937
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3025025214319115,
+      "learning_rate": 4.5382705864953915e-06,
+      "loss": 0.5308,
+      "step": 8938
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1459573447426146,
+      "learning_rate": 4.536165963612995e-06,
+      "loss": 0.5085,
+      "step": 8939
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1297496562230431,
+      "learning_rate": 4.534061685690633e-06,
+      "loss": 0.4663,
+      "step": 8940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1866729265397942,
+      "learning_rate": 4.531957752861161e-06,
+      "loss": 0.5364,
+      "step": 8941
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0903828152533144,
+      "learning_rate": 4.5298541652574016e-06,
+      "loss": 0.4813,
+      "step": 8942
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.130244481553155,
+      "learning_rate": 4.527750923012172e-06,
+      "loss": 0.5402,
+      "step": 8943
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2343152671557724,
+      "learning_rate": 4.525648026258264e-06,
+      "loss": 0.5803,
+      "step": 8944
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2548806750802566,
+      "learning_rate": 4.5235454751284355e-06,
+      "loss": 0.4966,
+      "step": 8945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2473058754399593,
+      "learning_rate": 4.52144326975544e-06,
+      "loss": 0.5163,
+      "step": 8946
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2336144337481356,
+      "learning_rate": 4.5193414102719935e-06,
+      "loss": 0.5444,
+      "step": 8947
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1298813881696133,
+      "learning_rate": 4.5172398968108e-06,
+      "loss": 0.4969,
+      "step": 8948
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1866582597656203,
+      "learning_rate": 4.515138729504539e-06,
+      "loss": 0.5538,
+      "step": 8949
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1886517811661692,
+      "learning_rate": 4.513037908485868e-06,
+      "loss": 0.496,
+      "step": 8950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1325533340223817,
+      "learning_rate": 4.5109374338874254e-06,
+      "loss": 0.5065,
+      "step": 8951
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1962072975086617,
+      "learning_rate": 4.508837305841821e-06,
+      "loss": 0.4837,
+      "step": 8952
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0876591160444244,
+      "learning_rate": 4.506737524481647e-06,
+      "loss": 0.4869,
+      "step": 8953
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.183129650818177,
+      "learning_rate": 4.504638089939478e-06,
+      "loss": 0.5118,
+      "step": 8954
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1534481123128846,
+      "learning_rate": 4.502539002347856e-06,
+      "loss": 0.4832,
+      "step": 8955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0818674672799473,
+      "learning_rate": 4.500440261839313e-06,
+      "loss": 0.5361,
+      "step": 8956
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1698920703078248,
+      "learning_rate": 4.498341868546347e-06,
+      "loss": 0.4734,
+      "step": 8957
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1398021917892027,
+      "learning_rate": 4.496243822601443e-06,
+      "loss": 0.5307,
+      "step": 8958
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.116735567035923,
+      "learning_rate": 4.494146124137062e-06,
+      "loss": 0.5238,
+      "step": 8959
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.103240014150979,
+      "learning_rate": 4.4920487732856425e-06,
+      "loss": 0.4728,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1926066870319436,
+      "learning_rate": 4.489951770179606e-06,
+      "loss": 0.5121,
+      "step": 8961
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2199523570185355,
+      "learning_rate": 4.487855114951337e-06,
+      "loss": 0.503,
+      "step": 8962
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.20279337288249,
+      "learning_rate": 4.485758807733215e-06,
+      "loss": 0.5601,
+      "step": 8963
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.368473298483761,
+      "learning_rate": 4.483662848657593e-06,
+      "loss": 0.5541,
+      "step": 8964
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.294734705405784,
+      "learning_rate": 4.481567237856792e-06,
+      "loss": 0.5298,
+      "step": 8965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.166117090254619,
+      "learning_rate": 4.479471975463125e-06,
+      "loss": 0.4939,
+      "step": 8966
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1123919230908959,
+      "learning_rate": 4.477377061608873e-06,
+      "loss": 0.5006,
+      "step": 8967
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2363939312962728,
+      "learning_rate": 4.475282496426298e-06,
+      "loss": 0.4813,
+      "step": 8968
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2289990079082709,
+      "learning_rate": 4.473188280047644e-06,
+      "loss": 0.5464,
+      "step": 8969
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2900006666107822,
+      "learning_rate": 4.471094412605128e-06,
+      "loss": 0.4877,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2796262360790707,
+      "learning_rate": 4.469000894230947e-06,
+      "loss": 0.5291,
+      "step": 8971
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.438104419519546,
+      "learning_rate": 4.46690772505728e-06,
+      "loss": 0.5639,
+      "step": 8972
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.143021456793221,
+      "learning_rate": 4.464814905216271e-06,
+      "loss": 0.48,
+      "step": 8973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2237088483490053,
+      "learning_rate": 4.46272243484006e-06,
+      "loss": 0.5276,
+      "step": 8974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1318541583749597,
+      "learning_rate": 4.4606303140607456e-06,
+      "loss": 0.514,
+      "step": 8975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2253948412242925,
+      "learning_rate": 4.458538543010418e-06,
+      "loss": 0.5543,
+      "step": 8976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.119878662042073,
+      "learning_rate": 4.456447121821147e-06,
+      "loss": 0.5048,
+      "step": 8977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1276534894180283,
+      "learning_rate": 4.454356050624966e-06,
+      "loss": 0.493,
+      "step": 8978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.234279966921267,
+      "learning_rate": 4.4522653295539e-06,
+      "loss": 0.5141,
+      "step": 8979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.161801799173484,
+      "learning_rate": 4.450174958739945e-06,
+      "loss": 0.4985,
+      "step": 8980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2165572663769564,
+      "learning_rate": 4.448084938315079e-06,
+      "loss": 0.5171,
+      "step": 8981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2023938917743457,
+      "learning_rate": 4.4459952684112596e-06,
+      "loss": 0.5307,
+      "step": 8982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1567017987070833,
+      "learning_rate": 4.443905949160409e-06,
+      "loss": 0.5185,
+      "step": 8983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1106023849048148,
+      "learning_rate": 4.441816980694446e-06,
+      "loss": 0.5237,
+      "step": 8984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1334221613424291,
+      "learning_rate": 4.439728363145248e-06,
+      "loss": 0.4591,
+      "step": 8985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1269532307801617,
+      "learning_rate": 4.437640096644687e-06,
+      "loss": 0.4717,
+      "step": 8986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.128895162244155,
+      "learning_rate": 4.435552181324608e-06,
+      "loss": 0.5089,
+      "step": 8987
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3376907087890968,
+      "learning_rate": 4.433464617316825e-06,
+      "loss": 0.5502,
+      "step": 8988
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2668222016076862,
+      "learning_rate": 4.431377404753141e-06,
+      "loss": 0.5537,
+      "step": 8989
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1287194331731123,
+      "learning_rate": 4.429290543765331e-06,
+      "loss": 0.4858,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.198636538055483,
+      "learning_rate": 4.42720403448515e-06,
+      "loss": 0.4949,
+      "step": 8991
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2931505519136397,
+      "learning_rate": 4.4251178770443334e-06,
+      "loss": 0.5091,
+      "step": 8992
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.286485210966762,
+      "learning_rate": 4.4230320715745855e-06,
+      "loss": 0.5279,
+      "step": 8993
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1815235426724866,
+      "learning_rate": 4.420946618207599e-06,
+      "loss": 0.5304,
+      "step": 8994
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0962344154807344,
+      "learning_rate": 4.418861517075034e-06,
+      "loss": 0.4803,
+      "step": 8995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.137327208337767,
+      "learning_rate": 4.416776768308535e-06,
+      "loss": 0.4889,
+      "step": 8996
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2428856093571454,
+      "learning_rate": 4.4146923720397285e-06,
+      "loss": 0.5146,
+      "step": 8997
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2433412100531351,
+      "learning_rate": 4.412608328400205e-06,
+      "loss": 0.5488,
+      "step": 8998
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1965893676004902,
+      "learning_rate": 4.410524637521545e-06,
+      "loss": 0.489,
+      "step": 8999
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1757382103022482,
+      "learning_rate": 4.408441299535302e-06,
+      "loss": 0.5419,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1486985824584184,
+      "learning_rate": 4.406358314573009e-06,
+      "loss": 0.525,
+      "step": 9001
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2258168280393857,
+      "learning_rate": 4.404275682766179e-06,
+      "loss": 0.5373,
+      "step": 9002
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.227782741932641,
+      "learning_rate": 4.402193404246291e-06,
+      "loss": 0.5119,
+      "step": 9003
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3380673925194606,
+      "learning_rate": 4.400111479144818e-06,
+      "loss": 0.5753,
+      "step": 9004
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3106562970981084,
+      "learning_rate": 4.398029907593197e-06,
+      "loss": 0.5248,
+      "step": 9005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.226884593100216,
+      "learning_rate": 4.395948689722847e-06,
+      "loss": 0.5076,
+      "step": 9006
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1029174263584895,
+      "learning_rate": 4.393867825665176e-06,
+      "loss": 0.4753,
+      "step": 9007
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2616214773751981,
+      "learning_rate": 4.391787315551548e-06,
+      "loss": 0.5051,
+      "step": 9008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2207651351437323,
+      "learning_rate": 4.389707159513321e-06,
+      "loss": 0.5314,
+      "step": 9009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2251088463829816,
+      "learning_rate": 4.387627357681827e-06,
+      "loss": 0.5467,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.343512536291729,
+      "learning_rate": 4.3855479101883735e-06,
+      "loss": 0.4916,
+      "step": 9011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.367318762881241,
+      "learning_rate": 4.383468817164251e-06,
+      "loss": 0.5804,
+      "step": 9012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1305766172162948,
+      "learning_rate": 4.3813900787407175e-06,
+      "loss": 0.5422,
+      "step": 9013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2445914561046298,
+      "learning_rate": 4.379311695049018e-06,
+      "loss": 0.5362,
+      "step": 9014
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1317631037319846,
+      "learning_rate": 4.377233666220368e-06,
+      "loss": 0.478,
+      "step": 9015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1509479409748156,
+      "learning_rate": 4.375155992385965e-06,
+      "loss": 0.5206,
+      "step": 9016
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2466173178209359,
+      "learning_rate": 4.373078673676988e-06,
+      "loss": 0.5729,
+      "step": 9017
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1610129946655845,
+      "learning_rate": 4.371001710224583e-06,
+      "loss": 0.5047,
+      "step": 9018
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3180130567265718,
+      "learning_rate": 4.368925102159881e-06,
+      "loss": 0.6043,
+      "step": 9019
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2063429267943142,
+      "learning_rate": 4.36684884961399e-06,
+      "loss": 0.5395,
+      "step": 9020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.063843830959475,
+      "learning_rate": 4.364772952717995e-06,
+      "loss": 0.4738,
+      "step": 9021
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2126084642957016,
+      "learning_rate": 4.362697411602961e-06,
+      "loss": 0.5186,
+      "step": 9022
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3467054468337225,
+      "learning_rate": 4.36062222639992e-06,
+      "loss": 0.5187,
+      "step": 9023
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.008991350422317,
+      "learning_rate": 4.358547397239896e-06,
+      "loss": 0.4546,
+      "step": 9024
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1955933583724399,
+      "learning_rate": 4.356472924253878e-06,
+      "loss": 0.4827,
+      "step": 9025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.167043466119726,
+      "learning_rate": 4.354398807572841e-06,
+      "loss": 0.5093,
+      "step": 9026
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1582693045687293,
+      "learning_rate": 4.352325047327735e-06,
+      "loss": 0.4977,
+      "step": 9027
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1859898752902895,
+      "learning_rate": 4.350251643649491e-06,
+      "loss": 0.5098,
+      "step": 9028
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1979494118015914,
+      "learning_rate": 4.348178596669006e-06,
+      "loss": 0.5576,
+      "step": 9029
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1964071908354001,
+      "learning_rate": 4.346105906517165e-06,
+      "loss": 0.5045,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0921139606390746,
+      "learning_rate": 4.344033573324829e-06,
+      "loss": 0.5427,
+      "step": 9031
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1642329456046006,
+      "learning_rate": 4.341961597222837e-06,
+      "loss": 0.517,
+      "step": 9032
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2031739955685403,
+      "learning_rate": 4.339889978341998e-06,
+      "loss": 0.5285,
+      "step": 9033
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1498781367708157,
+      "learning_rate": 4.337818716813112e-06,
+      "loss": 0.4556,
+      "step": 9034
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1543747412532068,
+      "learning_rate": 4.3357478127669376e-06,
+      "loss": 0.5595,
+      "step": 9035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1339921793995993,
+      "learning_rate": 4.333677266334228e-06,
+      "loss": 0.5066,
+      "step": 9036
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.143158072634017,
+      "learning_rate": 4.331607077645708e-06,
+      "loss": 0.5123,
+      "step": 9037
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.130208145821317,
+      "learning_rate": 4.329537246832081e-06,
+      "loss": 0.5337,
+      "step": 9038
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2730728282547898,
+      "learning_rate": 4.32746777402402e-06,
+      "loss": 0.569,
+      "step": 9039
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2290372726524488,
+      "learning_rate": 4.325398659352185e-06,
+      "loss": 0.5115,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1891656036390024,
+      "learning_rate": 4.32332990294721e-06,
+      "loss": 0.5049,
+      "step": 9041
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1179078987243145,
+      "learning_rate": 4.321261504939709e-06,
+      "loss": 0.4646,
+      "step": 9042
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1863923176972653,
+      "learning_rate": 4.319193465460264e-06,
+      "loss": 0.4984,
+      "step": 9043
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1185951168319976,
+      "learning_rate": 4.317125784639447e-06,
+      "loss": 0.5301,
+      "step": 9044
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.176268315106781,
+      "learning_rate": 4.315058462607798e-06,
+      "loss": 0.5121,
+      "step": 9045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2470900997781682,
+      "learning_rate": 4.312991499495838e-06,
+      "loss": 0.4866,
+      "step": 9046
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1629646808184575,
+      "learning_rate": 4.3109248954340644e-06,
+      "loss": 0.5193,
+      "step": 9047
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2061735395796462,
+      "learning_rate": 4.3088586505529584e-06,
+      "loss": 0.5191,
+      "step": 9048
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2681183920021164,
+      "learning_rate": 4.306792764982964e-06,
+      "loss": 0.5095,
+      "step": 9049
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2479141952838317,
+      "learning_rate": 4.304727238854517e-06,
+      "loss": 0.5393,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1450066734519282,
+      "learning_rate": 4.302662072298022e-06,
+      "loss": 0.5013,
+      "step": 9051
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0914050441508638,
+      "learning_rate": 4.300597265443869e-06,
+      "loss": 0.4534,
+      "step": 9052
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2837500488073235,
+      "learning_rate": 4.298532818422411e-06,
+      "loss": 0.5182,
+      "step": 9053
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2041853653579049,
+      "learning_rate": 4.296468731363996e-06,
+      "loss": 0.5444,
+      "step": 9054
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.21620430761527,
+      "learning_rate": 4.294405004398933e-06,
+      "loss": 0.5009,
+      "step": 9055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0903579975321351,
+      "learning_rate": 4.292341637657519e-06,
+      "loss": 0.4994,
+      "step": 9056
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2316055133144848,
+      "learning_rate": 4.290278631270025e-06,
+      "loss": 0.5261,
+      "step": 9057
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0141801974465687,
+      "learning_rate": 4.288215985366702e-06,
+      "loss": 0.4315,
+      "step": 9058
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0767958367420276,
+      "learning_rate": 4.286153700077771e-06,
+      "loss": 0.5066,
+      "step": 9059
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.177301988576125,
+      "learning_rate": 4.284091775533436e-06,
+      "loss": 0.5125,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.17649758655059,
+      "learning_rate": 4.282030211863876e-06,
+      "loss": 0.4898,
+      "step": 9061
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.182321550885035,
+      "learning_rate": 4.279969009199254e-06,
+      "loss": 0.5145,
+      "step": 9062
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.051888488755576,
+      "learning_rate": 4.277908167669696e-06,
+      "loss": 0.4722,
+      "step": 9063
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1668018137901053,
+      "learning_rate": 4.275847687405323e-06,
+      "loss": 0.4904,
+      "step": 9064
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2407152580271852,
+      "learning_rate": 4.273787568536212e-06,
+      "loss": 0.5195,
+      "step": 9065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1544091804206766,
+      "learning_rate": 4.271727811192437e-06,
+      "loss": 0.5086,
+      "step": 9066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2598565593694218,
+      "learning_rate": 4.269668415504038e-06,
+      "loss": 0.5537,
+      "step": 9067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2668573949086492,
+      "learning_rate": 4.2676093816010415e-06,
+      "loss": 0.5703,
+      "step": 9068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1430526921599626,
+      "learning_rate": 4.265550709613435e-06,
+      "loss": 0.507,
+      "step": 9069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.252846718803969,
+      "learning_rate": 4.263492399671198e-06,
+      "loss": 0.5123,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3555196703050318,
+      "learning_rate": 4.261434451904284e-06,
+      "loss": 0.5459,
+      "step": 9071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.229745028360689,
+      "learning_rate": 4.259376866442623e-06,
+      "loss": 0.4873,
+      "step": 9072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.255695955313812,
+      "learning_rate": 4.2573196434161135e-06,
+      "loss": 0.5207,
+      "step": 9073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1803718754360313,
+      "learning_rate": 4.255262782954648e-06,
+      "loss": 0.5578,
+      "step": 9074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2344349713786555,
+      "learning_rate": 4.253206285188079e-06,
+      "loss": 0.5368,
+      "step": 9075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4108657953876795,
+      "learning_rate": 4.251150150246245e-06,
+      "loss": 0.575,
+      "step": 9076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1229832374123374,
+      "learning_rate": 4.249094378258962e-06,
+      "loss": 0.4783,
+      "step": 9077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3340768628431836,
+      "learning_rate": 4.247038969356027e-06,
+      "loss": 0.4533,
+      "step": 9078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1259958839578421,
+      "learning_rate": 4.244983923667199e-06,
+      "loss": 0.4871,
+      "step": 9079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2534941474403605,
+      "learning_rate": 4.242929241322228e-06,
+      "loss": 0.5428,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.213416091686869,
+      "learning_rate": 4.2408749224508365e-06,
+      "loss": 0.5219,
+      "step": 9081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2258433280673946,
+      "learning_rate": 4.238820967182727e-06,
+      "loss": 0.5277,
+      "step": 9082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2000336503238085,
+      "learning_rate": 4.236767375647572e-06,
+      "loss": 0.5672,
+      "step": 9083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2368155869858863,
+      "learning_rate": 4.234714147975029e-06,
+      "loss": 0.4728,
+      "step": 9084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2055063337003815,
+      "learning_rate": 4.2326612842947225e-06,
+      "loss": 0.4983,
+      "step": 9085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1654620366044424,
+      "learning_rate": 4.230608784736267e-06,
+      "loss": 0.4837,
+      "step": 9086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.060445313482052,
+      "learning_rate": 4.228556649429243e-06,
+      "loss": 0.4678,
+      "step": 9087
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2776551406629741,
+      "learning_rate": 4.226504878503215e-06,
+      "loss": 0.5178,
+      "step": 9088
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.299215066965357,
+      "learning_rate": 4.224453472087725e-06,
+      "loss": 0.5927,
+      "step": 9089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2028548692915406,
+      "learning_rate": 4.2224024303122826e-06,
+      "loss": 0.5724,
+      "step": 9090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1625284929782964,
+      "learning_rate": 4.220351753306382e-06,
+      "loss": 0.5354,
+      "step": 9091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2092231674892457,
+      "learning_rate": 4.218301441199499e-06,
+      "loss": 0.498,
+      "step": 9092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2556249895186018,
+      "learning_rate": 4.216251494121071e-06,
+      "loss": 0.5134,
+      "step": 9093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.190240959246921,
+      "learning_rate": 4.2142019122005295e-06,
+      "loss": 0.5008,
+      "step": 9094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2247433627140112,
+      "learning_rate": 4.21215269556727e-06,
+      "loss": 0.5458,
+      "step": 9095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1163885821258495,
+      "learning_rate": 4.210103844350671e-06,
+      "loss": 0.5175,
+      "step": 9096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.226696031433189,
+      "learning_rate": 4.208055358680089e-06,
+      "loss": 0.5964,
+      "step": 9097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1163335884639174,
+      "learning_rate": 4.2060072386848535e-06,
+      "loss": 0.5178,
+      "step": 9098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1274637159781098,
+      "learning_rate": 4.20395948449428e-06,
+      "loss": 0.542,
+      "step": 9099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1440689663572732,
+      "learning_rate": 4.201912096237643e-06,
+      "loss": 0.5457,
+      "step": 9100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.169601422015864,
+      "learning_rate": 4.1998650740442096e-06,
+      "loss": 0.5159,
+      "step": 9101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.284009938207532,
+      "learning_rate": 4.197818418043221e-06,
+      "loss": 0.5242,
+      "step": 9102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.236423482727802,
+      "learning_rate": 4.19577212836389e-06,
+      "loss": 0.5448,
+      "step": 9103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.31498469448973,
+      "learning_rate": 4.193726205135412e-06,
+      "loss": 0.5721,
+      "step": 9104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2688654635652281,
+      "learning_rate": 4.191680648486952e-06,
+      "loss": 0.5855,
+      "step": 9105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2228991964848057,
+      "learning_rate": 4.18963545854766e-06,
+      "loss": 0.5339,
+      "step": 9106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2019483088265752,
+      "learning_rate": 4.187590635446659e-06,
+      "loss": 0.4596,
+      "step": 9107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2065018168663766,
+      "learning_rate": 4.185546179313049e-06,
+      "loss": 0.5293,
+      "step": 9108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1991801302279743,
+      "learning_rate": 4.183502090275911e-06,
+      "loss": 0.536,
+      "step": 9109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.115730402479264,
+      "learning_rate": 4.181458368464293e-06,
+      "loss": 0.5061,
+      "step": 9110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1519502646849555,
+      "learning_rate": 4.179415014007227e-06,
+      "loss": 0.5176,
+      "step": 9111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3001728181410255,
+      "learning_rate": 4.177372027033724e-06,
+      "loss": 0.5684,
+      "step": 9112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3123773335854758,
+      "learning_rate": 4.175329407672763e-06,
+      "loss": 0.5427,
+      "step": 9113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2562943768610535,
+      "learning_rate": 4.1732871560533105e-06,
+      "loss": 0.5222,
+      "step": 9114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1608950128392148,
+      "learning_rate": 4.1712452723043e-06,
+      "loss": 0.5034,
+      "step": 9115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1486048673782692,
+      "learning_rate": 4.169203756554646e-06,
+      "loss": 0.5158,
+      "step": 9116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2299901122184236,
+      "learning_rate": 4.167162608933243e-06,
+      "loss": 0.5481,
+      "step": 9117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2585822172956644,
+      "learning_rate": 4.1651218295689576e-06,
+      "loss": 0.5697,
+      "step": 9118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0749486489562008,
+      "learning_rate": 4.163081418590639e-06,
+      "loss": 0.4492,
+      "step": 9119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.214341185408686,
+      "learning_rate": 4.1610413761271005e-06,
+      "loss": 0.4727,
+      "step": 9120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.228963118472335,
+      "learning_rate": 4.159001702307146e-06,
+      "loss": 0.5474,
+      "step": 9121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1330237356091868,
+      "learning_rate": 4.156962397259553e-06,
+      "loss": 0.5053,
+      "step": 9122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2549235652733375,
+      "learning_rate": 4.154923461113066e-06,
+      "loss": 0.5257,
+      "step": 9123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2300608125245454,
+      "learning_rate": 4.152884893996421e-06,
+      "loss": 0.5583,
+      "step": 9124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4199510909777995,
+      "learning_rate": 4.1508466960383165e-06,
+      "loss": 0.5789,
+      "step": 9125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2658145432943335,
+      "learning_rate": 4.148808867367438e-06,
+      "loss": 0.5786,
+      "step": 9126
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.32282018935476,
+      "learning_rate": 4.146771408112443e-06,
+      "loss": 0.5474,
+      "step": 9127
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1603524912284828,
+      "learning_rate": 4.144734318401969e-06,
+      "loss": 0.4861,
+      "step": 9128
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3406056276002114,
+      "learning_rate": 4.14269759836463e-06,
+      "loss": 0.5396,
+      "step": 9129
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1821412090179717,
+      "learning_rate": 4.1406612481290066e-06,
+      "loss": 0.5314,
+      "step": 9130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1814413107502952,
+      "learning_rate": 4.138625267823669e-06,
+      "loss": 0.527,
+      "step": 9131
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1621023482422803,
+      "learning_rate": 4.136589657577164e-06,
+      "loss": 0.5375,
+      "step": 9132
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0597174054546443,
+      "learning_rate": 4.134554417518001e-06,
+      "loss": 0.4648,
+      "step": 9133
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2100368965450459,
+      "learning_rate": 4.132519547774678e-06,
+      "loss": 0.4732,
+      "step": 9134
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1344728601033054,
+      "learning_rate": 4.130485048475673e-06,
+      "loss": 0.4928,
+      "step": 9135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1242668624030703,
+      "learning_rate": 4.128450919749426e-06,
+      "loss": 0.5065,
+      "step": 9136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.183035841785023,
+      "learning_rate": 4.1264171617243655e-06,
+      "loss": 0.4838,
+      "step": 9137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.303411122112108,
+      "learning_rate": 4.124383774528893e-06,
+      "loss": 0.5825,
+      "step": 9138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0176304311360211,
+      "learning_rate": 4.122350758291387e-06,
+      "loss": 0.449,
+      "step": 9139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0302434112367764,
+      "learning_rate": 4.120318113140207e-06,
+      "loss": 0.4652,
+      "step": 9140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2378909581991357,
+      "learning_rate": 4.118285839203675e-06,
+      "loss": 0.5182,
+      "step": 9141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9856735021138948,
+      "learning_rate": 4.116253936610107e-06,
+      "loss": 0.453,
+      "step": 9142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2048638893752568,
+      "learning_rate": 4.114222405487781e-06,
+      "loss": 0.4992,
+      "step": 9143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.297629217501277,
+      "learning_rate": 4.112191245964962e-06,
+      "loss": 0.5592,
+      "step": 9144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.156058527583534,
+      "learning_rate": 4.110160458169888e-06,
+      "loss": 0.4844,
+      "step": 9145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3217858214541571,
+      "learning_rate": 4.10813004223077e-06,
+      "loss": 0.5343,
+      "step": 9146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.253360379913028,
+      "learning_rate": 4.106099998275801e-06,
+      "loss": 0.5238,
+      "step": 9147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2262744929893845,
+      "learning_rate": 4.104070326433146e-06,
+      "loss": 0.5616,
+      "step": 9148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2276122345279603,
+      "learning_rate": 4.102041026830952e-06,
+      "loss": 0.5136,
+      "step": 9149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.212738027509996,
+      "learning_rate": 4.100012099597339e-06,
+      "loss": 0.5255,
+      "step": 9150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1425299740288044,
+      "learning_rate": 4.0979835448604e-06,
+      "loss": 0.5023,
+      "step": 9151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3176413608263036,
+      "learning_rate": 4.095955362748214e-06,
+      "loss": 0.5592,
+      "step": 9152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.127117178097935,
+      "learning_rate": 4.093927553388822e-06,
+      "loss": 0.5445,
+      "step": 9153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1204352488501264,
+      "learning_rate": 4.091900116910256e-06,
+      "loss": 0.4846,
+      "step": 9154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1651226055315875,
+      "learning_rate": 4.089873053440521e-06,
+      "loss": 0.4943,
+      "step": 9155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2361898480506883,
+      "learning_rate": 4.087846363107588e-06,
+      "loss": 0.5198,
+      "step": 9156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1158977613207686,
+      "learning_rate": 4.085820046039417e-06,
+      "loss": 0.5175,
+      "step": 9157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2050143677895648,
+      "learning_rate": 4.083794102363939e-06,
+      "loss": 0.5262,
+      "step": 9158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1567126714496982,
+      "learning_rate": 4.081768532209064e-06,
+      "loss": 0.5194,
+      "step": 9159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2458617852490823,
+      "learning_rate": 4.079743335702679e-06,
+      "loss": 0.5365,
+      "step": 9160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2723189514059556,
+      "learning_rate": 4.077718512972638e-06,
+      "loss": 0.5193,
+      "step": 9161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1843739572485206,
+      "learning_rate": 4.075694064146786e-06,
+      "loss": 0.5017,
+      "step": 9162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.090513126352009,
+      "learning_rate": 4.07366998935293e-06,
+      "loss": 0.5133,
+      "step": 9163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1978434277688532,
+      "learning_rate": 4.071646288718863e-06,
+      "loss": 0.5584,
+      "step": 9164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.280004733568618,
+      "learning_rate": 4.069622962372355e-06,
+      "loss": 0.5757,
+      "step": 9165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.147629849602213,
+      "learning_rate": 4.067600010441143e-06,
+      "loss": 0.5448,
+      "step": 9166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1799589438488978,
+      "learning_rate": 4.06557743305295e-06,
+      "loss": 0.5137,
+      "step": 9167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.066421508679898,
+      "learning_rate": 4.06355523033547e-06,
+      "loss": 0.4865,
+      "step": 9168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2605736322107632,
+      "learning_rate": 4.0615334024163775e-06,
+      "loss": 0.5355,
+      "step": 9169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3299224359080555,
+      "learning_rate": 4.059511949423322e-06,
+      "loss": 0.5626,
+      "step": 9170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1347075823385895,
+      "learning_rate": 4.0574908714839245e-06,
+      "loss": 0.4856,
+      "step": 9171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2916612727555072,
+      "learning_rate": 4.05547016872579e-06,
+      "loss": 0.5258,
+      "step": 9172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2433949486104487,
+      "learning_rate": 4.05344984127649e-06,
+      "loss": 0.5682,
+      "step": 9173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1087062593108779,
+      "learning_rate": 4.051429889263582e-06,
+      "loss": 0.5214,
+      "step": 9174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1999989807601414,
+      "learning_rate": 4.049410312814598e-06,
+      "loss": 0.5506,
+      "step": 9175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1401144478887495,
+      "learning_rate": 4.0473911120570396e-06,
+      "loss": 0.5099,
+      "step": 9176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1448725689436217,
+      "learning_rate": 4.045372287118391e-06,
+      "loss": 0.4827,
+      "step": 9177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2162708106617781,
+      "learning_rate": 4.043353838126113e-06,
+      "loss": 0.533,
+      "step": 9178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1807670453691057,
+      "learning_rate": 4.041335765207638e-06,
+      "loss": 0.4799,
+      "step": 9179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1492767381592752,
+      "learning_rate": 4.039318068490383e-06,
+      "loss": 0.5103,
+      "step": 9180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1679734960351158,
+      "learning_rate": 4.037300748101728e-06,
+      "loss": 0.5272,
+      "step": 9181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1251872754517631,
+      "learning_rate": 4.0352838041690435e-06,
+      "loss": 0.52,
+      "step": 9182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0471188844736952,
+      "learning_rate": 4.033267236819664e-06,
+      "loss": 0.4527,
+      "step": 9183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2601790348648976,
+      "learning_rate": 4.031251046180906e-06,
+      "loss": 0.5334,
+      "step": 9184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1409762116141424,
+      "learning_rate": 4.029235232380069e-06,
+      "loss": 0.4469,
+      "step": 9185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2055794092143792,
+      "learning_rate": 4.027219795544413e-06,
+      "loss": 0.5227,
+      "step": 9186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1711845398384093,
+      "learning_rate": 4.025204735801187e-06,
+      "loss": 0.5013,
+      "step": 9187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2251611953867116,
+      "learning_rate": 4.023190053277612e-06,
+      "loss": 0.5534,
+      "step": 9188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3760132957304065,
+      "learning_rate": 4.021175748100885e-06,
+      "loss": 0.546,
+      "step": 9189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2826929108683494,
+      "learning_rate": 4.019161820398183e-06,
+      "loss": 0.5515,
+      "step": 9190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2477387001536355,
+      "learning_rate": 4.017148270296652e-06,
+      "loss": 0.5109,
+      "step": 9191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1340588257751902,
+      "learning_rate": 4.015135097923416e-06,
+      "loss": 0.4834,
+      "step": 9192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1923391723145023,
+      "learning_rate": 4.013122303405579e-06,
+      "loss": 0.5053,
+      "step": 9193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0859296510261816,
+      "learning_rate": 4.011109886870218e-06,
+      "loss": 0.4916,
+      "step": 9194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.188481577703572,
+      "learning_rate": 4.009097848444389e-06,
+      "loss": 0.4899,
+      "step": 9195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2902014122683947,
+      "learning_rate": 4.007086188255125e-06,
+      "loss": 0.5312,
+      "step": 9196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.113974104660785,
+      "learning_rate": 4.005074906429426e-06,
+      "loss": 0.4828,
+      "step": 9197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.149717331437313,
+      "learning_rate": 4.003064003094278e-06,
+      "loss": 0.5427,
+      "step": 9198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2601682980476445,
+      "learning_rate": 4.00105347837664e-06,
+      "loss": 0.4983,
+      "step": 9199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2905802721304305,
+      "learning_rate": 3.99904333240345e-06,
+      "loss": 0.5543,
+      "step": 9200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.235720757964405,
+      "learning_rate": 3.9970335653016146e-06,
+      "loss": 0.5093,
+      "step": 9201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1723006937589986,
+      "learning_rate": 3.995024177198018e-06,
+      "loss": 0.5109,
+      "step": 9202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2080630241241732,
+      "learning_rate": 3.993015168219527e-06,
+      "loss": 0.5561,
+      "step": 9203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1754927677336082,
+      "learning_rate": 3.991006538492981e-06,
+      "loss": 0.5244,
+      "step": 9204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0976940433607125,
+      "learning_rate": 3.9889982881451924e-06,
+      "loss": 0.4569,
+      "step": 9205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2162140603160938,
+      "learning_rate": 3.98699041730296e-06,
+      "loss": 0.5393,
+      "step": 9206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.292645094085734,
+      "learning_rate": 3.9849829260930416e-06,
+      "loss": 0.5385,
+      "step": 9207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.175047382454505,
+      "learning_rate": 3.982975814642185e-06,
+      "loss": 0.4959,
+      "step": 9208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1190184414991386,
+      "learning_rate": 3.980969083077108e-06,
+      "loss": 0.4977,
+      "step": 9209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.242665756574865,
+      "learning_rate": 3.9789627315245115e-06,
+      "loss": 0.5725,
+      "step": 9210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2050901440186441,
+      "learning_rate": 3.976956760111061e-06,
+      "loss": 0.4904,
+      "step": 9211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2056426916015859,
+      "learning_rate": 3.9749511689634025e-06,
+      "loss": 0.4892,
+      "step": 9212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.13164091363705,
+      "learning_rate": 3.972945958208162e-06,
+      "loss": 0.5123,
+      "step": 9213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.222067840656587,
+      "learning_rate": 3.9709411279719375e-06,
+      "loss": 0.5824,
+      "step": 9214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.193208575671176,
+      "learning_rate": 3.968936678381307e-06,
+      "loss": 0.4893,
+      "step": 9215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1840454347598417,
+      "learning_rate": 3.966932609562822e-06,
+      "loss": 0.5458,
+      "step": 9216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1467082700186808,
+      "learning_rate": 3.964928921643006e-06,
+      "loss": 0.4631,
+      "step": 9217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2393138442963556,
+      "learning_rate": 3.962925614748363e-06,
+      "loss": 0.5464,
+      "step": 9218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3700323098960099,
+      "learning_rate": 3.960922689005373e-06,
+      "loss": 0.4768,
+      "step": 9219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1401455537833538,
+      "learning_rate": 3.958920144540496e-06,
+      "loss": 0.4996,
+      "step": 9220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1672620900129118,
+      "learning_rate": 3.956917981480156e-06,
+      "loss": 0.5197,
+      "step": 9221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3190673735491731,
+      "learning_rate": 3.95491619995076e-06,
+      "loss": 0.5718,
+      "step": 9222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1202619171428119,
+      "learning_rate": 3.952914800078693e-06,
+      "loss": 0.4645,
+      "step": 9223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0950828605375833,
+      "learning_rate": 3.950913781990313e-06,
+      "loss": 0.4546,
+      "step": 9224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1518040831950365,
+      "learning_rate": 3.948913145811956e-06,
+      "loss": 0.461,
+      "step": 9225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2472318516288499,
+      "learning_rate": 3.946912891669934e-06,
+      "loss": 0.5452,
+      "step": 9226
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2215676138110405,
+      "learning_rate": 3.9449130196905275e-06,
+      "loss": 0.5369,
+      "step": 9227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3032043158454456,
+      "learning_rate": 3.942913530000002e-06,
+      "loss": 0.555,
+      "step": 9228
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1938240407780623,
+      "learning_rate": 3.940914422724597e-06,
+      "loss": 0.5054,
+      "step": 9229
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1565339538794939,
+      "learning_rate": 3.938915697990528e-06,
+      "loss": 0.5335,
+      "step": 9230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1029229387009631,
+      "learning_rate": 3.936917355923982e-06,
+      "loss": 0.5061,
+      "step": 9231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8393031660460972,
+      "learning_rate": 3.934919396651121e-06,
+      "loss": 0.4768,
+      "step": 9232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0909862477478638,
+      "learning_rate": 3.932921820298091e-06,
+      "loss": 0.4682,
+      "step": 9233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1714377541215373,
+      "learning_rate": 3.930924626991008e-06,
+      "loss": 0.5307,
+      "step": 9234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0746758060867339,
+      "learning_rate": 3.9289278168559665e-06,
+      "loss": 0.4797,
+      "step": 9235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.224787356935798,
+      "learning_rate": 3.926931390019036e-06,
+      "loss": 0.524,
+      "step": 9236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2695473537890891,
+      "learning_rate": 3.9249353466062575e-06,
+      "loss": 0.479,
+      "step": 9237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0773588581976077,
+      "learning_rate": 3.922939686743655e-06,
+      "loss": 0.4738,
+      "step": 9238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1997496085399793,
+      "learning_rate": 3.920944410557222e-06,
+      "loss": 0.489,
+      "step": 9239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3874839713055724,
+      "learning_rate": 3.918949518172936e-06,
+      "loss": 0.5511,
+      "step": 9240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2583054234186295,
+      "learning_rate": 3.916955009716741e-06,
+      "loss": 0.5154,
+      "step": 9241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2227689063482987,
+      "learning_rate": 3.914960885314557e-06,
+      "loss": 0.5448,
+      "step": 9242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.16173659043508,
+      "learning_rate": 3.912967145092288e-06,
+      "loss": 0.4664,
+      "step": 9243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1343483871827158,
+      "learning_rate": 3.910973789175807e-06,
+      "loss": 0.4785,
+      "step": 9244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2999481190819453,
+      "learning_rate": 3.908980817690966e-06,
+      "loss": 0.5031,
+      "step": 9245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.166828990043494,
+      "learning_rate": 3.906988230763592e-06,
+      "loss": 0.5191,
+      "step": 9246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2320738964646103,
+      "learning_rate": 3.90499602851949e-06,
+      "loss": 0.5247,
+      "step": 9247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0787309243116863,
+      "learning_rate": 3.903004211084431e-06,
+      "loss": 0.4801,
+      "step": 9248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2106633029721567,
+      "learning_rate": 3.901012778584172e-06,
+      "loss": 0.5072,
+      "step": 9249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.07588100483352,
+      "learning_rate": 3.8990217311444475e-06,
+      "loss": 0.4942,
+      "step": 9250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.087377677264296,
+      "learning_rate": 3.897031068890954e-06,
+      "loss": 0.4642,
+      "step": 9251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2158418829662845,
+      "learning_rate": 3.89504079194938e-06,
+      "loss": 0.5453,
+      "step": 9252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2622034423304875,
+      "learning_rate": 3.893050900445375e-06,
+      "loss": 0.5472,
+      "step": 9253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.170307178765335,
+      "learning_rate": 3.891061394504575e-06,
+      "loss": 0.4958,
+      "step": 9254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2740703952480115,
+      "learning_rate": 3.889072274252586e-06,
+      "loss": 0.5372,
+      "step": 9255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0767860390907344,
+      "learning_rate": 3.887083539814993e-06,
+      "loss": 0.462,
+      "step": 9256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1228507492638438,
+      "learning_rate": 3.885095191317357e-06,
+      "loss": 0.5027,
+      "step": 9257
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.275574558425168,
+      "learning_rate": 3.883107228885209e-06,
+      "loss": 0.5603,
+      "step": 9258
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2504615885109909,
+      "learning_rate": 3.881119652644059e-06,
+      "loss": 0.5159,
+      "step": 9259
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2307179032223476,
+      "learning_rate": 3.8791324627193996e-06,
+      "loss": 0.5248,
+      "step": 9260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2152022050641655,
+      "learning_rate": 3.877145659236682e-06,
+      "loss": 0.5123,
+      "step": 9261
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0877914520480494,
+      "learning_rate": 3.875159242321353e-06,
+      "loss": 0.4731,
+      "step": 9262
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.258201728448149,
+      "learning_rate": 3.873173212098818e-06,
+      "loss": 0.5634,
+      "step": 9263
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1698818295460138,
+      "learning_rate": 3.871187568694468e-06,
+      "loss": 0.5068,
+      "step": 9264
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.157321330507798,
+      "learning_rate": 3.869202312233668e-06,
+      "loss": 0.4622,
+      "step": 9265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1319227210042582,
+      "learning_rate": 3.8672174428417555e-06,
+      "loss": 0.4614,
+      "step": 9266
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.148809100385579,
+      "learning_rate": 3.865232960644051e-06,
+      "loss": 0.4694,
+      "step": 9267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1638486524944238,
+      "learning_rate": 3.8632488657658375e-06,
+      "loss": 0.5079,
+      "step": 9268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2258697789002149,
+      "learning_rate": 3.861265158332383e-06,
+      "loss": 0.508,
+      "step": 9269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2231359546231242,
+      "learning_rate": 3.859281838468937e-06,
+      "loss": 0.494,
+      "step": 9270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0851564411468009,
+      "learning_rate": 3.857298906300705e-06,
+      "loss": 0.5212,
+      "step": 9271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1181771756531849,
+      "learning_rate": 3.85531636195289e-06,
+      "loss": 0.4744,
+      "step": 9272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.174684995672424,
+      "learning_rate": 3.8533342055506505e-06,
+      "loss": 0.5124,
+      "step": 9273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2072320172393856,
+      "learning_rate": 3.851352437219137e-06,
+      "loss": 0.5216,
+      "step": 9274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2233539577134949,
+      "learning_rate": 3.849371057083465e-06,
+      "loss": 0.4894,
+      "step": 9275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0967186558255064,
+      "learning_rate": 3.8473900652687336e-06,
+      "loss": 0.4987,
+      "step": 9276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1696979900748865,
+      "learning_rate": 3.845409461900012e-06,
+      "loss": 0.523,
+      "step": 9277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2535209658025441,
+      "learning_rate": 3.843429247102343e-06,
+      "loss": 0.5261,
+      "step": 9278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2239009383792616,
+      "learning_rate": 3.841449421000748e-06,
+      "loss": 0.5455,
+      "step": 9279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2086710731908556,
+      "learning_rate": 3.839469983720229e-06,
+      "loss": 0.5539,
+      "step": 9280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2962919556201942,
+      "learning_rate": 3.837490935385751e-06,
+      "loss": 0.521,
+      "step": 9281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.156273609642032,
+      "learning_rate": 3.835512276122267e-06,
+      "loss": 0.4906,
+      "step": 9282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.156572348535469,
+      "learning_rate": 3.833534006054694e-06,
+      "loss": 0.5028,
+      "step": 9283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.282507442207382,
+      "learning_rate": 3.8315561253079344e-06,
+      "loss": 0.5582,
+      "step": 9284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1559097717375064,
+      "learning_rate": 3.829578634006862e-06,
+      "loss": 0.4893,
+      "step": 9285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3159027042105154,
+      "learning_rate": 3.827601532276325e-06,
+      "loss": 0.5633,
+      "step": 9286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0551041521847013,
+      "learning_rate": 3.825624820241153e-06,
+      "loss": 0.471,
+      "step": 9287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2604334281038085,
+      "learning_rate": 3.823648498026138e-06,
+      "loss": 0.5181,
+      "step": 9288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0864020801487757,
+      "learning_rate": 3.821672565756058e-06,
+      "loss": 0.4641,
+      "step": 9289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2304104745779894,
+      "learning_rate": 3.81969702355567e-06,
+      "loss": 0.5294,
+      "step": 9290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1659333728808234,
+      "learning_rate": 3.8177218715496915e-06,
+      "loss": 0.5302,
+      "step": 9291
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2711768181215732,
+      "learning_rate": 3.8157471098628295e-06,
+      "loss": 0.5153,
+      "step": 9292
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.210046895986327,
+      "learning_rate": 3.8137727386197564e-06,
+      "loss": 0.4897,
+      "step": 9293
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.110144496052831,
+      "learning_rate": 3.8117987579451275e-06,
+      "loss": 0.5018,
+      "step": 9294
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1120689885879511,
+      "learning_rate": 3.809825167963569e-06,
+      "loss": 0.4739,
+      "step": 9295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2287830249776177,
+      "learning_rate": 3.807851968799685e-06,
+      "loss": 0.5394,
+      "step": 9296
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.255823019162969,
+      "learning_rate": 3.8058791605780577e-06,
+      "loss": 0.5251,
+      "step": 9297
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2025825962242997,
+      "learning_rate": 3.8039067434232324e-06,
+      "loss": 0.5203,
+      "step": 9298
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.182422625084933,
+      "learning_rate": 3.8019347174597454e-06,
+      "loss": 0.5131,
+      "step": 9299
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1508865194348052,
+      "learning_rate": 3.7999630828120947e-06,
+      "loss": 0.4717,
+      "step": 9300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2077238694374473,
+      "learning_rate": 3.797991839604762e-06,
+      "loss": 0.4728,
+      "step": 9301
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1981971191287568,
+      "learning_rate": 3.7960209879622025e-06,
+      "loss": 0.5582,
+      "step": 9302
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2085117285131313,
+      "learning_rate": 3.79405052800885e-06,
+      "loss": 0.5187,
+      "step": 9303
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2688084819074847,
+      "learning_rate": 3.792080459869103e-06,
+      "loss": 0.5462,
+      "step": 9304
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.157197615506893,
+      "learning_rate": 3.7901107836673444e-06,
+      "loss": 0.4585,
+      "step": 9305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3535944047380466,
+      "learning_rate": 3.788141499527932e-06,
+      "loss": 0.5848,
+      "step": 9306
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2523252793994752,
+      "learning_rate": 3.7861726075751948e-06,
+      "loss": 0.5391,
+      "step": 9307
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.130220644601081,
+      "learning_rate": 3.7842041079334446e-06,
+      "loss": 0.5103,
+      "step": 9308
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1395874007092592,
+      "learning_rate": 3.7822360007269564e-06,
+      "loss": 0.4914,
+      "step": 9309
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.225183720338488,
+      "learning_rate": 3.780268286079988e-06,
+      "loss": 0.5236,
+      "step": 9310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2772257339134647,
+      "learning_rate": 3.77830096411677e-06,
+      "loss": 0.5128,
+      "step": 9311
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1542111540849496,
+      "learning_rate": 3.776334034961513e-06,
+      "loss": 0.4983,
+      "step": 9312
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.269094350904054,
+      "learning_rate": 3.7743674987384017e-06,
+      "loss": 0.5637,
+      "step": 9313
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1873309366443683,
+      "learning_rate": 3.7724013555715867e-06,
+      "loss": 0.4959,
+      "step": 9314
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1918699769112768,
+      "learning_rate": 3.7704356055852043e-06,
+      "loss": 0.4748,
+      "step": 9315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.225826018022242,
+      "learning_rate": 3.768470248903362e-06,
+      "loss": 0.549,
+      "step": 9316
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2046000863309725,
+      "learning_rate": 3.766505285650144e-06,
+      "loss": 0.5535,
+      "step": 9317
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3875861235687084,
+      "learning_rate": 3.7645407159496104e-06,
+      "loss": 0.4658,
+      "step": 9318
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1329643936925746,
+      "learning_rate": 3.762576539925793e-06,
+      "loss": 0.4972,
+      "step": 9319
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1519205641777288,
+      "learning_rate": 3.7606127577026965e-06,
+      "loss": 0.4955,
+      "step": 9320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2569953206809021,
+      "learning_rate": 3.7586493694043068e-06,
+      "loss": 0.5415,
+      "step": 9321
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1525568878134829,
+      "learning_rate": 3.7566863751545833e-06,
+      "loss": 0.5547,
+      "step": 9322
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.26928328219844,
+      "learning_rate": 3.7547237750774647e-06,
+      "loss": 0.6111,
+      "step": 9323
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2869445046862316,
+      "learning_rate": 3.7527615692968513e-06,
+      "loss": 0.5336,
+      "step": 9324
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2728515868503767,
+      "learning_rate": 3.7507997579366317e-06,
+      "loss": 0.5836,
+      "step": 9325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2223000296959026,
+      "learning_rate": 3.7488383411206654e-06,
+      "loss": 0.4893,
+      "step": 9326
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1165511446358587,
+      "learning_rate": 3.7468773189727857e-06,
+      "loss": 0.4623,
+      "step": 9327
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2386484173370365,
+      "learning_rate": 3.744916691616807e-06,
+      "loss": 0.551,
+      "step": 9328
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1764325844011927,
+      "learning_rate": 3.742956459176508e-06,
+      "loss": 0.5148,
+      "step": 9329
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2103421624387882,
+      "learning_rate": 3.7409966217756477e-06,
+      "loss": 0.4758,
+      "step": 9330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1721515583496773,
+      "learning_rate": 3.739037179537962e-06,
+      "loss": 0.5186,
+      "step": 9331
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2036209075252815,
+      "learning_rate": 3.737078132587163e-06,
+      "loss": 0.4936,
+      "step": 9332
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2664356225929865,
+      "learning_rate": 3.735119481046936e-06,
+      "loss": 0.5838,
+      "step": 9333
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1488216043265833,
+      "learning_rate": 3.7331612250409354e-06,
+      "loss": 0.5309,
+      "step": 9334
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2473215971686313,
+      "learning_rate": 3.7312033646928004e-06,
+      "loss": 0.5032,
+      "step": 9335
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1328500675680737,
+      "learning_rate": 3.7292459001261383e-06,
+      "loss": 0.5307,
+      "step": 9336
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1376216488021684,
+      "learning_rate": 3.7272888314645363e-06,
+      "loss": 0.4674,
+      "step": 9337
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0640978859102415,
+      "learning_rate": 3.725332158831556e-06,
+      "loss": 0.4742,
+      "step": 9338
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1563694350958473,
+      "learning_rate": 3.7233758823507303e-06,
+      "loss": 0.5684,
+      "step": 9339
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2073625523468285,
+      "learning_rate": 3.7214200021455647e-06,
+      "loss": 0.5513,
+      "step": 9340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0869069334826291,
+      "learning_rate": 3.719464518339547e-06,
+      "loss": 0.4856,
+      "step": 9341
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.206105936173928,
+      "learning_rate": 3.7175094310561375e-06,
+      "loss": 0.5413,
+      "step": 9342
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1111378699630299,
+      "learning_rate": 3.7155547404187754e-06,
+      "loss": 0.498,
+      "step": 9343
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.177880273388775,
+      "learning_rate": 3.7136004465508624e-06,
+      "loss": 0.5172,
+      "step": 9344
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2648978322502944,
+      "learning_rate": 3.711646549575786e-06,
+      "loss": 0.5064,
+      "step": 9345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.184756623638608,
+      "learning_rate": 3.709693049616907e-06,
+      "loss": 0.546,
+      "step": 9346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1899467659110678,
+      "learning_rate": 3.7077399467975594e-06,
+      "loss": 0.5459,
+      "step": 9347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0921243302791952,
+      "learning_rate": 3.7057872412410566e-06,
+      "loss": 0.5071,
+      "step": 9348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1971435484638682,
+      "learning_rate": 3.703834933070679e-06,
+      "loss": 0.5243,
+      "step": 9349
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1763220775707472,
+      "learning_rate": 3.7018830224096824e-06,
+      "loss": 0.5178,
+      "step": 9350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.102764582988765,
+      "learning_rate": 3.6999315093813048e-06,
+      "loss": 0.475,
+      "step": 9351
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2368553929560284,
+      "learning_rate": 3.6979803941087546e-06,
+      "loss": 0.5676,
+      "step": 9352
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1846536353887762,
+      "learning_rate": 3.696029676715216e-06,
+      "loss": 0.5264,
+      "step": 9353
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.159076072220522,
+      "learning_rate": 3.694079357323853e-06,
+      "loss": 0.5438,
+      "step": 9354
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1696867794164079,
+      "learning_rate": 3.69212943605779e-06,
+      "loss": 0.5435,
+      "step": 9355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3939657052937475,
+      "learning_rate": 3.6901799130401393e-06,
+      "loss": 0.5959,
+      "step": 9356
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1136429048675198,
+      "learning_rate": 3.688230788393986e-06,
+      "loss": 0.5107,
+      "step": 9357
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1645852265771934,
+      "learning_rate": 3.6862820622423913e-06,
+      "loss": 0.4796,
+      "step": 9358
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2764804267413663,
+      "learning_rate": 3.684333734708384e-06,
+      "loss": 0.5308,
+      "step": 9359
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.157024845618142,
+      "learning_rate": 3.68238580591497e-06,
+      "loss": 0.4817,
+      "step": 9360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.194352506209174,
+      "learning_rate": 3.680438275985133e-06,
+      "loss": 0.536,
+      "step": 9361
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2729384959961452,
+      "learning_rate": 3.6784911450418337e-06,
+      "loss": 0.5242,
+      "step": 9362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2697319928428525,
+      "learning_rate": 3.676544413208002e-06,
+      "loss": 0.5193,
+      "step": 9363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1208878703598866,
+      "learning_rate": 3.6745980806065507e-06,
+      "loss": 0.5217,
+      "step": 9364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2064323544404632,
+      "learning_rate": 3.6726521473603525e-06,
+      "loss": 0.5193,
+      "step": 9365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1178337308868647,
+      "learning_rate": 3.670706613592271e-06,
+      "loss": 0.4714,
+      "step": 9366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2075579336733664,
+      "learning_rate": 3.6687614794251348e-06,
+      "loss": 0.4973,
+      "step": 9367
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2252114989204077,
+      "learning_rate": 3.6668167449817548e-06,
+      "loss": 0.5328,
+      "step": 9368
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0726376636781376,
+      "learning_rate": 3.6648724103849086e-06,
+      "loss": 0.5084,
+      "step": 9369
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2401961193041842,
+      "learning_rate": 3.662928475757348e-06,
+      "loss": 0.54,
+      "step": 9370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.196323989046907,
+      "learning_rate": 3.6609849412218092e-06,
+      "loss": 0.5113,
+      "step": 9371
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.068363894046801,
+      "learning_rate": 3.6590418069009947e-06,
+      "loss": 0.5007,
+      "step": 9372
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1908152130453231,
+      "learning_rate": 3.657099072917587e-06,
+      "loss": 0.4856,
+      "step": 9373
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.080447418377442,
+      "learning_rate": 3.6551567393942422e-06,
+      "loss": 0.5003,
+      "step": 9374
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3184209964620741,
+      "learning_rate": 3.6532148064535855e-06,
+      "loss": 0.5416,
+      "step": 9375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1106599699835906,
+      "learning_rate": 3.6512732742182223e-06,
+      "loss": 0.4943,
+      "step": 9376
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2009596722491962,
+      "learning_rate": 3.649332142810732e-06,
+      "loss": 0.5597,
+      "step": 9377
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1328043970278625,
+      "learning_rate": 3.6473914123536725e-06,
+      "loss": 0.4907,
+      "step": 9378
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1424279269100082,
+      "learning_rate": 3.645451082969569e-06,
+      "loss": 0.4784,
+      "step": 9379
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.282054323015519,
+      "learning_rate": 3.64351115478092e-06,
+      "loss": 0.5729,
+      "step": 9380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1911864030959958,
+      "learning_rate": 3.6415716279102065e-06,
+      "loss": 0.4772,
+      "step": 9381
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2475869729832347,
+      "learning_rate": 3.6396325024798817e-06,
+      "loss": 0.5345,
+      "step": 9382
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1954470789327094,
+      "learning_rate": 3.6376937786123722e-06,
+      "loss": 0.4915,
+      "step": 9383
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1322935823192142,
+      "learning_rate": 3.6357554564300824e-06,
+      "loss": 0.4622,
+      "step": 9384
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1771800696308359,
+      "learning_rate": 3.6338175360553827e-06,
+      "loss": 0.59,
+      "step": 9385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1472558929996926,
+      "learning_rate": 3.6318800176106283e-06,
+      "loss": 0.524,
+      "step": 9386
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2529121808013093,
+      "learning_rate": 3.629942901218142e-06,
+      "loss": 0.5223,
+      "step": 9387
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1593039107827388,
+      "learning_rate": 3.6280061870002303e-06,
+      "loss": 0.5269,
+      "step": 9388
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1656866328331474,
+      "learning_rate": 3.6260698750791624e-06,
+      "loss": 0.4952,
+      "step": 9389
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.239932045920407,
+      "learning_rate": 3.6241339655771844e-06,
+      "loss": 0.5095,
+      "step": 9390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.131180264763077,
+      "learning_rate": 3.6221984586165247e-06,
+      "loss": 0.4894,
+      "step": 9391
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1253704944466592,
+      "learning_rate": 3.620263354319382e-06,
+      "loss": 0.5021,
+      "step": 9392
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2463284932080216,
+      "learning_rate": 3.6183286528079287e-06,
+      "loss": 0.5387,
+      "step": 9393
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1575249524650533,
+      "learning_rate": 3.6163943542043156e-06,
+      "loss": 0.464,
+      "step": 9394
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2383408396729982,
+      "learning_rate": 3.61446045863066e-06,
+      "loss": 0.5171,
+      "step": 9395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.238486769314854,
+      "learning_rate": 3.612526966209059e-06,
+      "loss": 0.529,
+      "step": 9396
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2254701840879783,
+      "learning_rate": 3.6105938770615902e-06,
+      "loss": 0.5012,
+      "step": 9397
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1532174112150761,
+      "learning_rate": 3.608661191310291e-06,
+      "loss": 0.4944,
+      "step": 9398
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2398912811302558,
+      "learning_rate": 3.6067289090771883e-06,
+      "loss": 0.5186,
+      "step": 9399
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2043771544409003,
+      "learning_rate": 3.6047970304842727e-06,
+      "loss": 0.5218,
+      "step": 9400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.207110405459133,
+      "learning_rate": 3.6028655556535142e-06,
+      "loss": 0.4893,
+      "step": 9401
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.232905756876433,
+      "learning_rate": 3.600934484706858e-06,
+      "loss": 0.506,
+      "step": 9402
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.203818579569143,
+      "learning_rate": 3.5990038177662234e-06,
+      "loss": 0.5199,
+      "step": 9403
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1479331225942453,
+      "learning_rate": 3.5970735549535065e-06,
+      "loss": 0.4979,
+      "step": 9404
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.278866293278656,
+      "learning_rate": 3.5951436963905663e-06,
+      "loss": 0.5025,
+      "step": 9405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2012205036523376,
+      "learning_rate": 3.59321424219925e-06,
+      "loss": 0.5143,
+      "step": 9406
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2108788260734868,
+      "learning_rate": 3.591285192501376e-06,
+      "loss": 0.5444,
+      "step": 9407
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1544723764603242,
+      "learning_rate": 3.58935654741873e-06,
+      "loss": 0.4352,
+      "step": 9408
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.148496509515144,
+      "learning_rate": 3.5874283070730787e-06,
+      "loss": 0.4986,
+      "step": 9409
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1626005272827842,
+      "learning_rate": 3.585500471586166e-06,
+      "loss": 0.4759,
+      "step": 9410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2613769635207994,
+      "learning_rate": 3.583573041079701e-06,
+      "loss": 0.5354,
+      "step": 9411
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2848791038105478,
+      "learning_rate": 3.581646015675374e-06,
+      "loss": 0.5735,
+      "step": 9412
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2028704782746054,
+      "learning_rate": 3.579719395494847e-06,
+      "loss": 0.5234,
+      "step": 9413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.266301056942487,
+      "learning_rate": 3.577793180659761e-06,
+      "loss": 0.5607,
+      "step": 9414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.194764504072646,
+      "learning_rate": 3.575867371291728e-06,
+      "loss": 0.5197,
+      "step": 9415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2372307879532023,
+      "learning_rate": 3.5739419675123275e-06,
+      "loss": 0.5199,
+      "step": 9416
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2479295750089943,
+      "learning_rate": 3.5720169694431294e-06,
+      "loss": 0.5336,
+      "step": 9417
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1309637835173993,
+      "learning_rate": 3.5700923772056606e-06,
+      "loss": 0.5064,
+      "step": 9418
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1952385973056323,
+      "learning_rate": 3.5681681909214338e-06,
+      "loss": 0.5145,
+      "step": 9419
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1421187039594531,
+      "learning_rate": 3.5662444107119365e-06,
+      "loss": 0.5244,
+      "step": 9420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2065019650749584,
+      "learning_rate": 3.5643210366986205e-06,
+      "loss": 0.4596,
+      "step": 9421
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.217841616804464,
+      "learning_rate": 3.5623980690029202e-06,
+      "loss": 0.5398,
+      "step": 9422
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.185905238983062,
+      "learning_rate": 3.560475507746244e-06,
+      "loss": 0.4767,
+      "step": 9423
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2575858723497246,
+      "learning_rate": 3.5585533530499726e-06,
+      "loss": 0.585,
+      "step": 9424
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1313473928267659,
+      "learning_rate": 3.556631605035464e-06,
+      "loss": 0.4843,
+      "step": 9425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1414290429149618,
+      "learning_rate": 3.554710263824043e-06,
+      "loss": 0.5394,
+      "step": 9426
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2171740490778613,
+      "learning_rate": 3.5527893295370196e-06,
+      "loss": 0.5175,
+      "step": 9427
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1237341328689883,
+      "learning_rate": 3.550868802295666e-06,
+      "loss": 0.4835,
+      "step": 9428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2128348468234456,
+      "learning_rate": 3.548948682221238e-06,
+      "loss": 0.5069,
+      "step": 9429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1712848957770694,
+      "learning_rate": 3.547028969434966e-06,
+      "loss": 0.5175,
+      "step": 9430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1700183147235923,
+      "learning_rate": 3.545109664058044e-06,
+      "loss": 0.4947,
+      "step": 9431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1560020696463091,
+      "learning_rate": 3.5431907662116528e-06,
+      "loss": 0.4864,
+      "step": 9432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1559592732082835,
+      "learning_rate": 3.5412722760169403e-06,
+      "loss": 0.4897,
+      "step": 9433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2422307804652986,
+      "learning_rate": 3.5393541935950327e-06,
+      "loss": 0.4787,
+      "step": 9434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1150600744751,
+      "learning_rate": 3.53743651906703e-06,
+      "loss": 0.4936,
+      "step": 9435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2099919226668443,
+      "learning_rate": 3.5355192525539996e-06,
+      "loss": 0.5095,
+      "step": 9436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1187619938553792,
+      "learning_rate": 3.5336023941769947e-06,
+      "loss": 0.4575,
+      "step": 9437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.186138376304257,
+      "learning_rate": 3.5316859440570284e-06,
+      "loss": 0.5227,
+      "step": 9438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2782591359979427,
+      "learning_rate": 3.5297699023151013e-06,
+      "loss": 0.5308,
+      "step": 9439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1651090487413547,
+      "learning_rate": 3.527854269072186e-06,
+      "loss": 0.5115,
+      "step": 9440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3675091174447633,
+      "learning_rate": 3.525939044449218e-06,
+      "loss": 0.5431,
+      "step": 9441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.346361284464671,
+      "learning_rate": 3.524024228567121e-06,
+      "loss": 0.5604,
+      "step": 9442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3062517759320174,
+      "learning_rate": 3.5221098215467852e-06,
+      "loss": 0.561,
+      "step": 9443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0930372368035426,
+      "learning_rate": 3.520195823509078e-06,
+      "loss": 0.4577,
+      "step": 9444
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2473986259965422,
+      "learning_rate": 3.518282234574845e-06,
+      "loss": 0.5526,
+      "step": 9445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1495726288844041,
+      "learning_rate": 3.5163690548648898e-06,
+      "loss": 0.4712,
+      "step": 9446
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2018343455500333,
+      "learning_rate": 3.514456284500012e-06,
+      "loss": 0.5224,
+      "step": 9447
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2390626674041059,
+      "learning_rate": 3.5125439236009674e-06,
+      "loss": 0.4882,
+      "step": 9448
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2288681034053315,
+      "learning_rate": 3.510631972288494e-06,
+      "loss": 0.5326,
+      "step": 9449
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.220903059798556,
+      "learning_rate": 3.508720430683309e-06,
+      "loss": 0.5283,
+      "step": 9450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.289878263421879,
+      "learning_rate": 3.5068092989060907e-06,
+      "loss": 0.5671,
+      "step": 9451
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2561371349287989,
+      "learning_rate": 3.504898577077502e-06,
+      "loss": 0.5717,
+      "step": 9452
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1130103065993753,
+      "learning_rate": 3.502988265318176e-06,
+      "loss": 0.5147,
+      "step": 9453
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1680463682174476,
+      "learning_rate": 3.501078363748721e-06,
+      "loss": 0.5658,
+      "step": 9454
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1959511042076074,
+      "learning_rate": 3.4991688724897223e-06,
+      "loss": 0.5251,
+      "step": 9455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2124657616254628,
+      "learning_rate": 3.49725979166173e-06,
+      "loss": 0.5339,
+      "step": 9456
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2430410271807268,
+      "learning_rate": 3.4953511213852785e-06,
+      "loss": 0.5359,
+      "step": 9457
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2150701085024065,
+      "learning_rate": 3.493442861780868e-06,
+      "loss": 0.5163,
+      "step": 9458
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.179041755224835,
+      "learning_rate": 3.4915350129689798e-06,
+      "loss": 0.5008,
+      "step": 9459
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2507684253080418,
+      "learning_rate": 3.489627575070068e-06,
+      "loss": 0.5181,
+      "step": 9460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.153933860123063,
+      "learning_rate": 3.487720548204553e-06,
+      "loss": 0.4959,
+      "step": 9461
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1792388972523369,
+      "learning_rate": 3.4858139324928388e-06,
+      "loss": 0.5336,
+      "step": 9462
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2206944335628083,
+      "learning_rate": 3.4839077280553e-06,
+      "loss": 0.5073,
+      "step": 9463
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0341639531048983,
+      "learning_rate": 3.4820019350122847e-06,
+      "loss": 0.462,
+      "step": 9464
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1788765349286616,
+      "learning_rate": 3.4800965534841158e-06,
+      "loss": 0.538,
+      "step": 9465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1516457722963354,
+      "learning_rate": 3.4781915835910927e-06,
+      "loss": 0.4807,
+      "step": 9466
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1233892565729153,
+      "learning_rate": 3.476287025453484e-06,
+      "loss": 0.5268,
+      "step": 9467
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2700723287306177,
+      "learning_rate": 3.474382879191529e-06,
+      "loss": 0.5311,
+      "step": 9468
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0271704349893325,
+      "learning_rate": 3.472479144925451e-06,
+      "loss": 0.4595,
+      "step": 9469
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3595503770670667,
+      "learning_rate": 3.4705758227754426e-06,
+      "loss": 0.5599,
+      "step": 9470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.125043974122935,
+      "learning_rate": 3.4686729128616726e-06,
+      "loss": 0.5038,
+      "step": 9471
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0696727204924688,
+      "learning_rate": 3.4667704153042758e-06,
+      "loss": 0.4675,
+      "step": 9472
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1897369697532185,
+      "learning_rate": 3.46486833022337e-06,
+      "loss": 0.5165,
+      "step": 9473
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.167925728608404,
+      "learning_rate": 3.462966657739042e-06,
+      "loss": 0.5232,
+      "step": 9474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.187137699071726,
+      "learning_rate": 3.461065397971357e-06,
+      "loss": 0.4585,
+      "step": 9475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2564532594742266,
+      "learning_rate": 3.4591645510403528e-06,
+      "loss": 0.6012,
+      "step": 9476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.213434954141269,
+      "learning_rate": 3.457264117066037e-06,
+      "loss": 0.4945,
+      "step": 9477
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0629648986683833,
+      "learning_rate": 3.45536409616839e-06,
+      "loss": 0.4273,
+      "step": 9478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2954013218813911,
+      "learning_rate": 3.453464488467373e-06,
+      "loss": 0.5355,
+      "step": 9479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1615639825277282,
+      "learning_rate": 3.4515652940829192e-06,
+      "loss": 0.452,
+      "step": 9480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1676074333507098,
+      "learning_rate": 3.449666513134937e-06,
+      "loss": 0.4498,
+      "step": 9481
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1809245312098,
+      "learning_rate": 3.4477681457433e-06,
+      "loss": 0.5256,
+      "step": 9482
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2467951698848505,
+      "learning_rate": 3.4458701920278646e-06,
+      "loss": 0.5342,
+      "step": 9483
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1513420276950452,
+      "learning_rate": 3.4439726521084595e-06,
+      "loss": 0.5352,
+      "step": 9484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2094466846127767,
+      "learning_rate": 3.4420755261048843e-06,
+      "loss": 0.4968,
+      "step": 9485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2738476894154822,
+      "learning_rate": 3.4401788141369196e-06,
+      "loss": 0.5073,
+      "step": 9486
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.28192293126791,
+      "learning_rate": 3.4382825163243106e-06,
+      "loss": 0.5609,
+      "step": 9487
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0975383545901471,
+      "learning_rate": 3.4363866327867768e-06,
+      "loss": 0.5158,
+      "step": 9488
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.360212221697563,
+      "learning_rate": 3.434491163644019e-06,
+      "loss": 0.5538,
+      "step": 9489
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0967503946937076,
+      "learning_rate": 3.432596109015708e-06,
+      "loss": 0.5207,
+      "step": 9490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2211018879937154,
+      "learning_rate": 3.430701469021491e-06,
+      "loss": 0.5025,
+      "step": 9491
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1892160263902438,
+      "learning_rate": 3.4288072437809794e-06,
+      "loss": 0.498,
+      "step": 9492
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2119828664966084,
+      "learning_rate": 3.4269134334137698e-06,
+      "loss": 0.5662,
+      "step": 9493
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2544732163794503,
+      "learning_rate": 3.4250200380394284e-06,
+      "loss": 0.5173,
+      "step": 9494
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2514662725862404,
+      "learning_rate": 3.4231270577774976e-06,
+      "loss": 0.5239,
+      "step": 9495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1575979160064678,
+      "learning_rate": 3.421234492747484e-06,
+      "loss": 0.5165,
+      "step": 9496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.305615078243322,
+      "learning_rate": 3.419342343068882e-06,
+      "loss": 0.5665,
+      "step": 9497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2087566796071885,
+      "learning_rate": 3.417450608861147e-06,
+      "loss": 0.4663,
+      "step": 9498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1930274814020905,
+      "learning_rate": 3.4155592902437162e-06,
+      "loss": 0.5326,
+      "step": 9499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2406117743644756,
+      "learning_rate": 3.4136683873359987e-06,
+      "loss": 0.5467,
+      "step": 9500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2319065473553306,
+      "learning_rate": 3.4117779002573803e-06,
+      "loss": 0.4889,
+      "step": 9501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2642127269600925,
+      "learning_rate": 3.40988782912721e-06,
+      "loss": 0.5367,
+      "step": 9502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2914795842611744,
+      "learning_rate": 3.4079981740648215e-06,
+      "loss": 0.5291,
+      "step": 9503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2091311364153738,
+      "learning_rate": 3.406108935189519e-06,
+      "loss": 0.5409,
+      "step": 9504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.199552160381181,
+      "learning_rate": 3.404220112620583e-06,
+      "loss": 0.446,
+      "step": 9505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2489497541059247,
+      "learning_rate": 3.402331706477258e-06,
+      "loss": 0.5278,
+      "step": 9506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.151139124316079,
+      "learning_rate": 3.400443716878774e-06,
+      "loss": 0.4736,
+      "step": 9507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2149511455399715,
+      "learning_rate": 3.398556143944325e-06,
+      "loss": 0.5619,
+      "step": 9508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2118997503844922,
+      "learning_rate": 3.3966689877930857e-06,
+      "loss": 0.5061,
+      "step": 9509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1959842464819288,
+      "learning_rate": 3.394782248544202e-06,
+      "loss": 0.5532,
+      "step": 9510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2025265877494704,
+      "learning_rate": 3.392895926316795e-06,
+      "loss": 0.5236,
+      "step": 9511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0676761653365265,
+      "learning_rate": 3.3910100212299547e-06,
+      "loss": 0.4885,
+      "step": 9512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.189237227365878,
+      "learning_rate": 3.3891245334027487e-06,
+      "loss": 0.5269,
+      "step": 9513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3206982472331639,
+      "learning_rate": 3.387239462954219e-06,
+      "loss": 0.5313,
+      "step": 9514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2133309613892842,
+      "learning_rate": 3.385354810003383e-06,
+      "loss": 0.4849,
+      "step": 9515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2180806302503446,
+      "learning_rate": 3.383470574669222e-06,
+      "loss": 0.5528,
+      "step": 9516
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.217348221829534,
+      "learning_rate": 3.3815867570707028e-06,
+      "loss": 0.4838,
+      "step": 9517
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3387243638491388,
+      "learning_rate": 3.3797033573267546e-06,
+      "loss": 0.5698,
+      "step": 9518
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2024565982718625,
+      "learning_rate": 3.377820375556291e-06,
+      "loss": 0.5233,
+      "step": 9519
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2012118201088033,
+      "learning_rate": 3.3759378118781917e-06,
+      "loss": 0.5467,
+      "step": 9520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.265468587747006,
+      "learning_rate": 3.3740556664113145e-06,
+      "loss": 0.5555,
+      "step": 9521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1315449430277773,
+      "learning_rate": 3.372173939274492e-06,
+      "loss": 0.5347,
+      "step": 9522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2533361737408477,
+      "learning_rate": 3.3702926305865202e-06,
+      "loss": 0.5456,
+      "step": 9523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1186749886027787,
+      "learning_rate": 3.36841174046618e-06,
+      "loss": 0.4308,
+      "step": 9524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1082871166909192,
+      "learning_rate": 3.3665312690322238e-06,
+      "loss": 0.4934,
+      "step": 9525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2138847171809726,
+      "learning_rate": 3.3646512164033696e-06,
+      "loss": 0.5159,
+      "step": 9526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1476236690625292,
+      "learning_rate": 3.362771582698321e-06,
+      "loss": 0.4966,
+      "step": 9527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.263644941889011,
+      "learning_rate": 3.3608923680357432e-06,
+      "loss": 0.5217,
+      "step": 9528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1583409348016596,
+      "learning_rate": 3.359013572534283e-06,
+      "loss": 0.5223,
+      "step": 9529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1760657083617911,
+      "learning_rate": 3.3571351963125596e-06,
+      "loss": 0.5018,
+      "step": 9530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.207479005003192,
+      "learning_rate": 3.3552572394891635e-06,
+      "loss": 0.4748,
+      "step": 9531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.155805244480373,
+      "learning_rate": 3.353379702182664e-06,
+      "loss": 0.4969,
+      "step": 9532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1666799726181558,
+      "learning_rate": 3.3515025845115923e-06,
+      "loss": 0.4845,
+      "step": 9533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2978001707076032,
+      "learning_rate": 3.349625886594464e-06,
+      "loss": 0.51,
+      "step": 9534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2741658754681031,
+      "learning_rate": 3.3477496085497685e-06,
+      "loss": 0.5663,
+      "step": 9535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2649210632332923,
+      "learning_rate": 3.345873750495957e-06,
+      "loss": 0.5349,
+      "step": 9536
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2388800487906175,
+      "learning_rate": 3.343998312551471e-06,
+      "loss": 0.5229,
+      "step": 9537
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1441923297153567,
+      "learning_rate": 3.3421232948347084e-06,
+      "loss": 0.5154,
+      "step": 9538
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0550261907354799,
+      "learning_rate": 3.3402486974640522e-06,
+      "loss": 0.4497,
+      "step": 9539
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1793390732121094,
+      "learning_rate": 3.3383745205578555e-06,
+      "loss": 0.534,
+      "step": 9540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2234965596370864,
+      "learning_rate": 3.3365007642344447e-06,
+      "loss": 0.5412,
+      "step": 9541
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2635890455973273,
+      "learning_rate": 3.334627428612124e-06,
+      "loss": 0.4643,
+      "step": 9542
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2149688067572466,
+      "learning_rate": 3.3327545138091576e-06,
+      "loss": 0.5446,
+      "step": 9543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1730049788913282,
+      "learning_rate": 3.3308820199437987e-06,
+      "loss": 0.5009,
+      "step": 9544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2804875314858442,
+      "learning_rate": 3.3290099471342687e-06,
+      "loss": 0.5835,
+      "step": 9545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2764840222145815,
+      "learning_rate": 3.3271382954987554e-06,
+      "loss": 0.5901,
+      "step": 9546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1999281484232625,
+      "learning_rate": 3.325267065155431e-06,
+      "loss": 0.5276,
+      "step": 9547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1760959647623928,
+      "learning_rate": 3.323396256222432e-06,
+      "loss": 0.5212,
+      "step": 9548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1815716179488593,
+      "learning_rate": 3.321525868817873e-06,
+      "loss": 0.5038,
+      "step": 9549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3645097067191698,
+      "learning_rate": 3.319655903059843e-06,
+      "loss": 0.5188,
+      "step": 9550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2459354599055037,
+      "learning_rate": 3.3177863590664027e-06,
+      "loss": 0.5191,
+      "step": 9551
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3007752931017413,
+      "learning_rate": 3.315917236955587e-06,
+      "loss": 0.5706,
+      "step": 9552
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2057969779382451,
+      "learning_rate": 3.314048536845399e-06,
+      "loss": 0.537,
+      "step": 9553
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2124738729749147,
+      "learning_rate": 3.312180258853822e-06,
+      "loss": 0.5345,
+      "step": 9554
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0849620362421304,
+      "learning_rate": 3.3103124030988133e-06,
+      "loss": 0.4574,
+      "step": 9555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1479840064446605,
+      "learning_rate": 3.308444969698292e-06,
+      "loss": 0.5157,
+      "step": 9556
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1782256409741256,
+      "learning_rate": 3.3065779587701686e-06,
+      "loss": 0.5117,
+      "step": 9557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.195709461792309,
+      "learning_rate": 3.3047113704323085e-06,
+      "loss": 0.4774,
+      "step": 9558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.176182725869622,
+      "learning_rate": 3.302845204802563e-06,
+      "loss": 0.5136,
+      "step": 9559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3085574927929533,
+      "learning_rate": 3.300979461998751e-06,
+      "loss": 0.5196,
+      "step": 9560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2593984615895066,
+      "learning_rate": 3.2991141421386696e-06,
+      "loss": 0.5173,
+      "step": 9561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2253522794571168,
+      "learning_rate": 3.2972492453400873e-06,
+      "loss": 0.5662,
+      "step": 9562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2340296491557334,
+      "learning_rate": 3.2953847717207375e-06,
+      "loss": 0.5348,
+      "step": 9563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1331738125975577,
+      "learning_rate": 3.29352072139834e-06,
+      "loss": 0.5102,
+      "step": 9564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.257181281696201,
+      "learning_rate": 3.291657094490582e-06,
+      "loss": 0.5326,
+      "step": 9565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2155917891253631,
+      "learning_rate": 3.2897938911151196e-06,
+      "loss": 0.5237,
+      "step": 9566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.195306316683864,
+      "learning_rate": 3.287931111389593e-06,
+      "loss": 0.5074,
+      "step": 9567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.289391400542176,
+      "learning_rate": 3.2860687554316006e-06,
+      "loss": 0.5756,
+      "step": 9568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1225045926652168,
+      "learning_rate": 3.2842068233587275e-06,
+      "loss": 0.4464,
+      "step": 9569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0891565049701193,
+      "learning_rate": 3.2823453152885266e-06,
+      "loss": 0.4572,
+      "step": 9570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2095725457558757,
+      "learning_rate": 3.280484231338524e-06,
+      "loss": 0.5182,
+      "step": 9571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1536845020619162,
+      "learning_rate": 3.2786235716262236e-06,
+      "loss": 0.5242,
+      "step": 9572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2496519080910367,
+      "learning_rate": 3.276763336269092e-06,
+      "loss": 0.507,
+      "step": 9573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1976518863566135,
+      "learning_rate": 3.2749035253845773e-06,
+      "loss": 0.5128,
+      "step": 9574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3234836935113166,
+      "learning_rate": 3.273044139090105e-06,
+      "loss": 0.5381,
+      "step": 9575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2696282921895148,
+      "learning_rate": 3.271185177503058e-06,
+      "loss": 0.5333,
+      "step": 9576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1917294423479057,
+      "learning_rate": 3.2693266407408064e-06,
+      "loss": 0.4782,
+      "step": 9577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1496572439821997,
+      "learning_rate": 3.267468528920694e-06,
+      "loss": 0.5007,
+      "step": 9578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0720940680225253,
+      "learning_rate": 3.265610842160025e-06,
+      "loss": 0.4496,
+      "step": 9579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2297235563583229,
+      "learning_rate": 3.263753580576089e-06,
+      "loss": 0.5162,
+      "step": 9580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.224461450723111,
+      "learning_rate": 3.261896744286144e-06,
+      "loss": 0.498,
+      "step": 9581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1557453703922704,
+      "learning_rate": 3.2600403334074205e-06,
+      "loss": 0.5217,
+      "step": 9582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1602727144397882,
+      "learning_rate": 3.2581843480571285e-06,
+      "loss": 0.5032,
+      "step": 9583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2238037769319816,
+      "learning_rate": 3.2563287883524376e-06,
+      "loss": 0.5503,
+      "step": 9584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1477382373543588,
+      "learning_rate": 3.254473654410507e-06,
+      "loss": 0.5313,
+      "step": 9585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.227598882293441,
+      "learning_rate": 3.2526189463484536e-06,
+      "loss": 0.5105,
+      "step": 9586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2740430737703992,
+      "learning_rate": 3.250764664283378e-06,
+      "loss": 0.5357,
+      "step": 9587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1937142451658767,
+      "learning_rate": 3.248910808332354e-06,
+      "loss": 0.468,
+      "step": 9588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2481086250037112,
+      "learning_rate": 3.2470573786124184e-06,
+      "loss": 0.511,
+      "step": 9589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.140881914442385,
+      "learning_rate": 3.2452043752405914e-06,
+      "loss": 0.4756,
+      "step": 9590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1246766049428927,
+      "learning_rate": 3.2433517983338627e-06,
+      "loss": 0.4674,
+      "step": 9591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1718132511400805,
+      "learning_rate": 3.241499648009193e-06,
+      "loss": 0.5062,
+      "step": 9592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2260687253305598,
+      "learning_rate": 3.2396479243835243e-06,
+      "loss": 0.4923,
+      "step": 9593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3285709754295572,
+      "learning_rate": 3.237796627573757e-06,
+      "loss": 0.5492,
+      "step": 9594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2990738026868491,
+      "learning_rate": 3.2359457576967812e-06,
+      "loss": 0.5298,
+      "step": 9595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1824105772922286,
+      "learning_rate": 3.2340953148694444e-06,
+      "loss": 0.4678,
+      "step": 9596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1391189247253153,
+      "learning_rate": 3.2322452992085775e-06,
+      "loss": 0.4816,
+      "step": 9597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1966098402037928,
+      "learning_rate": 3.2303957108309846e-06,
+      "loss": 0.4823,
+      "step": 9598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1605868583989596,
+      "learning_rate": 3.2285465498534343e-06,
+      "loss": 0.5056,
+      "step": 9599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1648864901995553,
+      "learning_rate": 3.2266978163926765e-06,
+      "loss": 0.4967,
+      "step": 9600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.053097975296423,
+      "learning_rate": 3.2248495105654308e-06,
+      "loss": 0.4267,
+      "step": 9601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.269275299093629,
+      "learning_rate": 3.2230016324883906e-06,
+      "loss": 0.5859,
+      "step": 9602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.260502326303444,
+      "learning_rate": 3.2211541822782255e-06,
+      "loss": 0.5236,
+      "step": 9603
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2566710795992027,
+      "learning_rate": 3.2193071600515678e-06,
+      "loss": 0.5346,
+      "step": 9604
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1875216833443867,
+      "learning_rate": 3.2174605659250367e-06,
+      "loss": 0.5626,
+      "step": 9605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1233260628684765,
+      "learning_rate": 3.21561440001521e-06,
+      "loss": 0.4808,
+      "step": 9606
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2106343535936837,
+      "learning_rate": 3.213768662438649e-06,
+      "loss": 0.5286,
+      "step": 9607
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2317500149645764,
+      "learning_rate": 3.2119233533118864e-06,
+      "loss": 0.5129,
+      "step": 9608
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.273805857548278,
+      "learning_rate": 3.2100784727514235e-06,
+      "loss": 0.5209,
+      "step": 9609
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2752639216859851,
+      "learning_rate": 3.2082340208737363e-06,
+      "loss": 0.4705,
+      "step": 9610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2152096114644306,
+      "learning_rate": 3.206389997795277e-06,
+      "loss": 0.5098,
+      "step": 9611
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.139154557640684,
+      "learning_rate": 3.204546403632468e-06,
+      "loss": 0.4852,
+      "step": 9612
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2513026126009654,
+      "learning_rate": 3.2027032385017065e-06,
+      "loss": 0.5046,
+      "step": 9613
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2896581949537642,
+      "learning_rate": 3.2008605025193564e-06,
+      "loss": 0.5802,
+      "step": 9614
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2939197926433892,
+      "learning_rate": 3.199018195801765e-06,
+      "loss": 0.5221,
+      "step": 9615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.300807230921914,
+      "learning_rate": 3.1971763184652404e-06,
+      "loss": 0.527,
+      "step": 9616
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2733365352367334,
+      "learning_rate": 3.1953348706260723e-06,
+      "loss": 0.5559,
+      "step": 9617
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.218033066202077,
+      "learning_rate": 3.1934938524005243e-06,
+      "loss": 0.5444,
+      "step": 9618
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2230639280764783,
+      "learning_rate": 3.1916532639048237e-06,
+      "loss": 0.527,
+      "step": 9619
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.244658214273411,
+      "learning_rate": 3.1898131052551784e-06,
+      "loss": 0.5449,
+      "step": 9620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2124241716722137,
+      "learning_rate": 3.187973376567769e-06,
+      "loss": 0.4891,
+      "step": 9621
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2611029099287843,
+      "learning_rate": 3.1861340779587444e-06,
+      "loss": 0.5272,
+      "step": 9622
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.234839919213569,
+      "learning_rate": 3.1842952095442335e-06,
+      "loss": 0.5283,
+      "step": 9623
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.155577180325893,
+      "learning_rate": 3.182456771440329e-06,
+      "loss": 0.5187,
+      "step": 9624
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1982110974693967,
+      "learning_rate": 3.1806187637631035e-06,
+      "loss": 0.4872,
+      "step": 9625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3175504336032817,
+      "learning_rate": 3.178781186628597e-06,
+      "loss": 0.5188,
+      "step": 9626
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.132336957365549,
+      "learning_rate": 3.176944040152826e-06,
+      "loss": 0.4636,
+      "step": 9627
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1718656921017068,
+      "learning_rate": 3.1751073244517817e-06,
+      "loss": 0.4784,
+      "step": 9628
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2166551534987615,
+      "learning_rate": 3.1732710396414257e-06,
+      "loss": 0.5338,
+      "step": 9629
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.220696923812374,
+      "learning_rate": 3.1714351858376867e-06,
+      "loss": 0.498,
+      "step": 9630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2440496917502224,
+      "learning_rate": 3.1695997631564756e-06,
+      "loss": 0.5264,
+      "step": 9631
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2735555395170235,
+      "learning_rate": 3.167764771713673e-06,
+      "loss": 0.5668,
+      "step": 9632
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2009856785310227,
+      "learning_rate": 3.165930211625131e-06,
+      "loss": 0.4957,
+      "step": 9633
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1969525922120872,
+      "learning_rate": 3.1640960830066723e-06,
+      "loss": 0.5027,
+      "step": 9634
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.141580573314053,
+      "learning_rate": 3.1622623859740998e-06,
+      "loss": 0.4756,
+      "step": 9635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2348034272144184,
+      "learning_rate": 3.160429120643177e-06,
+      "loss": 0.5102,
+      "step": 9636
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1220948537827005,
+      "learning_rate": 3.1585962871296514e-06,
+      "loss": 0.5148,
+      "step": 9637
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1861523460012184,
+      "learning_rate": 3.15676388554924e-06,
+      "loss": 0.5513,
+      "step": 9638
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1555888889214319,
+      "learning_rate": 3.154931916017633e-06,
+      "loss": 0.5081,
+      "step": 9639
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2182523249262076,
+      "learning_rate": 3.1531003786504877e-06,
+      "loss": 0.5487,
+      "step": 9640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1503821795271751,
+      "learning_rate": 3.151269273563441e-06,
+      "loss": 0.4865,
+      "step": 9641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1729633617874444,
+      "learning_rate": 3.149438600872099e-06,
+      "loss": 0.5045,
+      "step": 9642
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0840679514515028,
+      "learning_rate": 3.147608360692046e-06,
+      "loss": 0.4947,
+      "step": 9643
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1614280947639186,
+      "learning_rate": 3.1457785531388275e-06,
+      "loss": 0.5187,
+      "step": 9644
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1710159713639752,
+      "learning_rate": 3.1439491783279754e-06,
+      "loss": 0.5399,
+      "step": 9645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2505672121589768,
+      "learning_rate": 3.1421202363749816e-06,
+      "loss": 0.5327,
+      "step": 9646
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.210515496338738,
+      "learning_rate": 3.1402917273953183e-06,
+      "loss": 0.554,
+      "step": 9647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2018282453832836,
+      "learning_rate": 3.13846365150443e-06,
+      "loss": 0.5085,
+      "step": 9648
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1599178390345357,
+      "learning_rate": 3.136636008817736e-06,
+      "loss": 0.5033,
+      "step": 9649
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1208079967447666,
+      "learning_rate": 3.134808799450617e-06,
+      "loss": 0.4635,
+      "step": 9650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1668448255383752,
+      "learning_rate": 3.1329820235184392e-06,
+      "loss": 0.5017,
+      "step": 9651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1973444304426357,
+      "learning_rate": 3.131155681136535e-06,
+      "loss": 0.5014,
+      "step": 9652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2872798268471883,
+      "learning_rate": 3.129329772420214e-06,
+      "loss": 0.573,
+      "step": 9653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1669371268718376,
+      "learning_rate": 3.12750429748475e-06,
+      "loss": 0.4799,
+      "step": 9654
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1191617154835367,
+      "learning_rate": 3.1256792564453997e-06,
+      "loss": 0.4965,
+      "step": 9655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1682877631195296,
+      "learning_rate": 3.1238546494173814e-06,
+      "loss": 0.4773,
+      "step": 9656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2555377364417983,
+      "learning_rate": 3.122030476515896e-06,
+      "loss": 0.5172,
+      "step": 9657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2261297835471494,
+      "learning_rate": 3.120206737856112e-06,
+      "loss": 0.4972,
+      "step": 9658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2282230774832554,
+      "learning_rate": 3.1183834335531748e-06,
+      "loss": 0.4974,
+      "step": 9659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2060289387908014,
+      "learning_rate": 3.116560563722193e-06,
+      "loss": 0.4786,
+      "step": 9660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1947679463531546,
+      "learning_rate": 3.1147381284782562e-06,
+      "loss": 0.545,
+      "step": 9661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4195915583058847,
+      "learning_rate": 3.112916127936425e-06,
+      "loss": 0.5533,
+      "step": 9662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1560191877779025,
+      "learning_rate": 3.111094562211735e-06,
+      "loss": 0.4798,
+      "step": 9663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1619429781738808,
+      "learning_rate": 3.109273431419183e-06,
+      "loss": 0.4808,
+      "step": 9664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1951582566018901,
+      "learning_rate": 3.107452735673755e-06,
+      "loss": 0.5223,
+      "step": 9665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.32950197532596,
+      "learning_rate": 3.1056324750903934e-06,
+      "loss": 0.5584,
+      "step": 9666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2145058017716204,
+      "learning_rate": 3.1038126497840225e-06,
+      "loss": 0.5298,
+      "step": 9667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1613504961298915,
+      "learning_rate": 3.1019932598695402e-06,
+      "loss": 0.4781,
+      "step": 9668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2345857561035964,
+      "learning_rate": 3.100174305461815e-06,
+      "loss": 0.4983,
+      "step": 9669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2016430930701383,
+      "learning_rate": 3.0983557866756818e-06,
+      "loss": 0.5158,
+      "step": 9670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1028421963427857,
+      "learning_rate": 3.096537703625955e-06,
+      "loss": 0.5089,
+      "step": 9671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2161646098852747,
+      "learning_rate": 3.0947200564274206e-06,
+      "loss": 0.4954,
+      "step": 9672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.351248864624224,
+      "learning_rate": 3.092902845194837e-06,
+      "loss": 0.5527,
+      "step": 9673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.076700790226702,
+      "learning_rate": 3.0910860700429315e-06,
+      "loss": 0.4599,
+      "step": 9674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1896682820272197,
+      "learning_rate": 3.0892697310864107e-06,
+      "loss": 0.4777,
+      "step": 9675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1848758514172655,
+      "learning_rate": 3.0874538284399424e-06,
+      "loss": 0.531,
+      "step": 9676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0745385820927151,
+      "learning_rate": 3.0856383622181785e-06,
+      "loss": 0.5203,
+      "step": 9677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1340142551313297,
+      "learning_rate": 3.083823332535738e-06,
+      "loss": 0.5117,
+      "step": 9678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2985881382932598,
+      "learning_rate": 3.0820087395072172e-06,
+      "loss": 0.5805,
+      "step": 9679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2501920552532362,
+      "learning_rate": 3.0801945832471736e-06,
+      "loss": 0.5804,
+      "step": 9680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2163620075272463,
+      "learning_rate": 3.078380863870146e-06,
+      "loss": 0.4998,
+      "step": 9681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.114146595652614,
+      "learning_rate": 3.076567581490647e-06,
+      "loss": 0.4944,
+      "step": 9682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2460728945526496,
+      "learning_rate": 3.0747547362231588e-06,
+      "loss": 0.5611,
+      "step": 9683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2827749249139446,
+      "learning_rate": 3.072942328182131e-06,
+      "loss": 0.54,
+      "step": 9684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2000758683539983,
+      "learning_rate": 3.071130357481996e-06,
+      "loss": 0.5028,
+      "step": 9685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2700625672284762,
+      "learning_rate": 3.0693188242371464e-06,
+      "loss": 0.5593,
+      "step": 9686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2973007973873976,
+      "learning_rate": 3.067507728561958e-06,
+      "loss": 0.4982,
+      "step": 9687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0810890364595844,
+      "learning_rate": 3.065697070570772e-06,
+      "loss": 0.4986,
+      "step": 9688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2541923315436387,
+      "learning_rate": 3.0638868503779075e-06,
+      "loss": 0.5283,
+      "step": 9689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2070078924690613,
+      "learning_rate": 3.0620770680976554e-06,
+      "loss": 0.536,
+      "step": 9690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1986306702506266,
+      "learning_rate": 3.060267723844269e-06,
+      "loss": 0.4594,
+      "step": 9691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2195843506028903,
+      "learning_rate": 3.058458817731985e-06,
+      "loss": 0.5094,
+      "step": 9692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2256512506719115,
+      "learning_rate": 3.0566503498750135e-06,
+      "loss": 0.5288,
+      "step": 9693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.198845373423006,
+      "learning_rate": 3.0548423203875245e-06,
+      "loss": 0.4933,
+      "step": 9694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.224672063412419,
+      "learning_rate": 3.0530347293836758e-06,
+      "loss": 0.4683,
+      "step": 9695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.269420020292581,
+      "learning_rate": 3.0512275769775834e-06,
+      "loss": 0.5721,
+      "step": 9696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1948019196331745,
+      "learning_rate": 3.0494208632833443e-06,
+      "loss": 0.5035,
+      "step": 9697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.156843548578526,
+      "learning_rate": 3.0476145884150265e-06,
+      "loss": 0.4974,
+      "step": 9698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2688441368470538,
+      "learning_rate": 3.0458087524866697e-06,
+      "loss": 0.5456,
+      "step": 9699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1903322976434654,
+      "learning_rate": 3.0440033556122883e-06,
+      "loss": 0.5462,
+      "step": 9700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0663380583594542,
+      "learning_rate": 3.04219839790586e-06,
+      "loss": 0.4863,
+      "step": 9701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.020107880852447,
+      "learning_rate": 3.0403938794813448e-06,
+      "loss": 0.4864,
+      "step": 9702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1512526179313922,
+      "learning_rate": 3.0385898004526725e-06,
+      "loss": 0.4166,
+      "step": 9703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1410519049236627,
+      "learning_rate": 3.0367861609337412e-06,
+      "loss": 0.4846,
+      "step": 9704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1613091286017547,
+      "learning_rate": 3.0349829610384274e-06,
+      "loss": 0.4816,
+      "step": 9705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.112162727493802,
+      "learning_rate": 3.03318020088057e-06,
+      "loss": 0.5055,
+      "step": 9706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1611980038141168,
+      "learning_rate": 3.031377880573991e-06,
+      "loss": 0.4837,
+      "step": 9707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1704282858681387,
+      "learning_rate": 3.0295760002324804e-06,
+      "loss": 0.5137,
+      "step": 9708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1833039486061843,
+      "learning_rate": 3.0277745599697996e-06,
+      "loss": 0.5305,
+      "step": 9709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1505943338181377,
+      "learning_rate": 3.025973559899685e-06,
+      "loss": 0.51,
+      "step": 9710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2589359360530683,
+      "learning_rate": 3.0241730001358383e-06,
+      "loss": 0.517,
+      "step": 9711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2993040202517023,
+      "learning_rate": 3.0223728807919406e-06,
+      "loss": 0.6323,
+      "step": 9712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2725565384780952,
+      "learning_rate": 3.020573201981646e-06,
+      "loss": 0.5138,
+      "step": 9713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2987660328643698,
+      "learning_rate": 3.018773963818571e-06,
+      "loss": 0.5545,
+      "step": 9714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2787221767944847,
+      "learning_rate": 3.016975166416317e-06,
+      "loss": 0.5264,
+      "step": 9715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2301707073318102,
+      "learning_rate": 3.015176809888445e-06,
+      "loss": 0.5101,
+      "step": 9716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.309659381656543,
+      "learning_rate": 3.0133788943484987e-06,
+      "loss": 0.505,
+      "step": 9717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1884990805982245,
+      "learning_rate": 3.011581419909988e-06,
+      "loss": 0.5427,
+      "step": 9718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1365239871204527,
+      "learning_rate": 3.0097843866863985e-06,
+      "loss": 0.5214,
+      "step": 9719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1672198596232213,
+      "learning_rate": 3.0079877947911883e-06,
+      "loss": 0.4931,
+      "step": 9720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2158417849195762,
+      "learning_rate": 3.0061916443377805e-06,
+      "loss": 0.4731,
+      "step": 9721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2957583756350022,
+      "learning_rate": 3.004395935439577e-06,
+      "loss": 0.5218,
+      "step": 9722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2692954445968987,
+      "learning_rate": 3.002600668209953e-06,
+      "loss": 0.5631,
+      "step": 9723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2135708140964503,
+      "learning_rate": 3.000805842762248e-06,
+      "loss": 0.5186,
+      "step": 9724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1757931629477214,
+      "learning_rate": 2.999011459209784e-06,
+      "loss": 0.4773,
+      "step": 9725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2630064448949543,
+      "learning_rate": 2.9972175176658448e-06,
+      "loss": 0.5139,
+      "step": 9726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2904038357530452,
+      "learning_rate": 2.995424018243692e-06,
+      "loss": 0.5303,
+      "step": 9727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0653556707110878,
+      "learning_rate": 2.9936309610565606e-06,
+      "loss": 0.4723,
+      "step": 9728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2497233084575268,
+      "learning_rate": 2.9918383462176547e-06,
+      "loss": 0.4913,
+      "step": 9729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2938495874084577,
+      "learning_rate": 2.9900461738401545e-06,
+      "loss": 0.5679,
+      "step": 9730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2638106351015859,
+      "learning_rate": 2.988254444037203e-06,
+      "loss": 0.519,
+      "step": 9731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.204300392979703,
+      "learning_rate": 2.9864631569219237e-06,
+      "loss": 0.4885,
+      "step": 9732
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1726332690829093,
+      "learning_rate": 2.984672312607414e-06,
+      "loss": 0.4703,
+      "step": 9733
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2400835561287917,
+      "learning_rate": 2.982881911206733e-06,
+      "loss": 0.5033,
+      "step": 9734
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.241585735696034,
+      "learning_rate": 2.981091952832923e-06,
+      "loss": 0.4944,
+      "step": 9735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2364526960099262,
+      "learning_rate": 2.9793024375989877e-06,
+      "loss": 0.5357,
+      "step": 9736
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2138513270943172,
+      "learning_rate": 2.9775133656179113e-06,
+      "loss": 0.5658,
+      "step": 9737
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2376171461202323,
+      "learning_rate": 2.975724737002648e-06,
+      "loss": 0.557,
+      "step": 9738
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1327460960130638,
+      "learning_rate": 2.973936551866121e-06,
+      "loss": 0.4955,
+      "step": 9739
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.077405219368041,
+      "learning_rate": 2.97214881032123e-06,
+      "loss": 0.4842,
+      "step": 9740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.423243440533533,
+      "learning_rate": 2.9703615124808484e-06,
+      "loss": 0.605,
+      "step": 9741
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0908098211782857,
+      "learning_rate": 2.9685746584578078e-06,
+      "loss": 0.4956,
+      "step": 9742
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2265318848500568,
+      "learning_rate": 2.966788248364929e-06,
+      "loss": 0.5175,
+      "step": 9743
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1097582839996938,
+      "learning_rate": 2.9650022823149925e-06,
+      "loss": 0.3929,
+      "step": 9744
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.162524596333507,
+      "learning_rate": 2.9632167604207586e-06,
+      "loss": 0.4391,
+      "step": 9745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2255626904231984,
+      "learning_rate": 2.9614316827949574e-06,
+      "loss": 0.4882,
+      "step": 9746
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1049093172138844,
+      "learning_rate": 2.959647049550286e-06,
+      "loss": 0.4703,
+      "step": 9747
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2790026610334093,
+      "learning_rate": 2.9578628607994187e-06,
+      "loss": 0.521,
+      "step": 9748
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2820677589922698,
+      "learning_rate": 2.956079116655003e-06,
+      "loss": 0.5336,
+      "step": 9749
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3268876089610115,
+      "learning_rate": 2.9542958172296533e-06,
+      "loss": 0.5115,
+      "step": 9750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1877541521048391,
+      "learning_rate": 2.9525129626359637e-06,
+      "loss": 0.5069,
+      "step": 9751
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2870209219239013,
+      "learning_rate": 2.950730552986487e-06,
+      "loss": 0.523,
+      "step": 9752
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.210638390793444,
+      "learning_rate": 2.948948588393764e-06,
+      "loss": 0.4904,
+      "step": 9753
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2198304986459967,
+      "learning_rate": 2.9471670689702927e-06,
+      "loss": 0.5251,
+      "step": 9754
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1133374116350807,
+      "learning_rate": 2.945385994828551e-06,
+      "loss": 0.4778,
+      "step": 9755
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.266781925721349,
+      "learning_rate": 2.9436053660809914e-06,
+      "loss": 0.5526,
+      "step": 9756
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.276616954225686,
+      "learning_rate": 2.941825182840029e-06,
+      "loss": 0.5367,
+      "step": 9757
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0861925882108026,
+      "learning_rate": 2.940045445218057e-06,
+      "loss": 0.4439,
+      "step": 9758
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.223251246748148,
+      "learning_rate": 2.9382661533274424e-06,
+      "loss": 0.5487,
+      "step": 9759
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2432689637472019,
+      "learning_rate": 2.936487307280518e-06,
+      "loss": 0.5079,
+      "step": 9760
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.341238735894706,
+      "learning_rate": 2.9347089071895963e-06,
+      "loss": 0.5597,
+      "step": 9761
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.199765804405271,
+      "learning_rate": 2.9329309531669505e-06,
+      "loss": 0.4765,
+      "step": 9762
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2067155641234408,
+      "learning_rate": 2.931153445324837e-06,
+      "loss": 0.529,
+      "step": 9763
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2344159952430254,
+      "learning_rate": 2.929376383775475e-06,
+      "loss": 0.5688,
+      "step": 9764
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1055547997510957,
+      "learning_rate": 2.927599768631061e-06,
+      "loss": 0.5002,
+      "step": 9765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.223891636539499,
+      "learning_rate": 2.9258236000037656e-06,
+      "loss": 0.5093,
+      "step": 9766
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0899147912938831,
+      "learning_rate": 2.9240478780057214e-06,
+      "loss": 0.5142,
+      "step": 9767
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2348902146195369,
+      "learning_rate": 2.9222726027490413e-06,
+      "loss": 0.53,
+      "step": 9768
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.276223674431669,
+      "learning_rate": 2.9204977743458084e-06,
+      "loss": 0.5536,
+      "step": 9769
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.21900057050834,
+      "learning_rate": 2.918723392908076e-06,
+      "loss": 0.5431,
+      "step": 9770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2101023101113764,
+      "learning_rate": 2.916949458547874e-06,
+      "loss": 0.5313,
+      "step": 9771
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1394050030882883,
+      "learning_rate": 2.9151759713771933e-06,
+      "loss": 0.5097,
+      "step": 9772
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2700705453848662,
+      "learning_rate": 2.9134029315080094e-06,
+      "loss": 0.5033,
+      "step": 9773
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2294417682275165,
+      "learning_rate": 2.911630339052257e-06,
+      "loss": 0.5199,
+      "step": 9774
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.14783249054286,
+      "learning_rate": 2.909858194121853e-06,
+      "loss": 0.5145,
+      "step": 9775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1075335518639304,
+      "learning_rate": 2.908086496828685e-06,
+      "loss": 0.4875,
+      "step": 9776
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1784986353766367,
+      "learning_rate": 2.906315247284602e-06,
+      "loss": 0.4948,
+      "step": 9777
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1317728994356206,
+      "learning_rate": 2.904544445601436e-06,
+      "loss": 0.482,
+      "step": 9778
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2665373257220667,
+      "learning_rate": 2.9027740918909873e-06,
+      "loss": 0.5572,
+      "step": 9779
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0476162691616189,
+      "learning_rate": 2.9010041862650273e-06,
+      "loss": 0.4903,
+      "step": 9780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1946357357017348,
+      "learning_rate": 2.899234728835302e-06,
+      "loss": 0.4706,
+      "step": 9781
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1590615704847174,
+      "learning_rate": 2.8974657197135203e-06,
+      "loss": 0.5094,
+      "step": 9782
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3795955329216594,
+      "learning_rate": 2.895697159011375e-06,
+      "loss": 0.534,
+      "step": 9783
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1630723569703318,
+      "learning_rate": 2.893929046840518e-06,
+      "loss": 0.4747,
+      "step": 9784
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1057504355705519,
+      "learning_rate": 2.892161383312583e-06,
+      "loss": 0.5357,
+      "step": 9785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1968714203544453,
+      "learning_rate": 2.8903941685391745e-06,
+      "loss": 0.4947,
+      "step": 9786
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2296795933237805,
+      "learning_rate": 2.8886274026318593e-06,
+      "loss": 0.5358,
+      "step": 9787
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2436282841152793,
+      "learning_rate": 2.886861085702186e-06,
+      "loss": 0.5366,
+      "step": 9788
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2671993965314863,
+      "learning_rate": 2.885095217861672e-06,
+      "loss": 0.5508,
+      "step": 9789
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2071877782368567,
+      "learning_rate": 2.8833297992218055e-06,
+      "loss": 0.5654,
+      "step": 9790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2480367501912277,
+      "learning_rate": 2.881564829894048e-06,
+      "loss": 0.5357,
+      "step": 9791
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2153232033126546,
+      "learning_rate": 2.8798003099898297e-06,
+      "loss": 0.4976,
+      "step": 9792
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1746961078901785,
+      "learning_rate": 2.8780362396205495e-06,
+      "loss": 0.4913,
+      "step": 9793
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0602920254125476,
+      "learning_rate": 2.8762726188975876e-06,
+      "loss": 0.4982,
+      "step": 9794
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1111228498527537,
+      "learning_rate": 2.874509447932288e-06,
+      "loss": 0.4899,
+      "step": 9795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1744743510596547,
+      "learning_rate": 2.872746726835969e-06,
+      "loss": 0.5362,
+      "step": 9796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2808508832858856,
+      "learning_rate": 2.870984455719924e-06,
+      "loss": 0.5181,
+      "step": 9797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2362708971901422,
+      "learning_rate": 2.8692226346954087e-06,
+      "loss": 0.4979,
+      "step": 9798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2511913820878486,
+      "learning_rate": 2.8674612638736576e-06,
+      "loss": 0.5314,
+      "step": 9799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3125190733477405,
+      "learning_rate": 2.865700343365877e-06,
+      "loss": 0.56,
+      "step": 9800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2053376201279824,
+      "learning_rate": 2.8639398732832448e-06,
+      "loss": 0.5058,
+      "step": 9801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3062222071707097,
+      "learning_rate": 2.862179853736905e-06,
+      "loss": 0.5553,
+      "step": 9802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1668391043521174,
+      "learning_rate": 2.860420284837975e-06,
+      "loss": 0.5037,
+      "step": 9803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2755488112114801,
+      "learning_rate": 2.858661166697547e-06,
+      "loss": 0.5874,
+      "step": 9804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2417344040710159,
+      "learning_rate": 2.8569024994266848e-06,
+      "loss": 0.5099,
+      "step": 9805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0850638298737119,
+      "learning_rate": 2.855144283136421e-06,
+      "loss": 0.491,
+      "step": 9806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1009858698618142,
+      "learning_rate": 2.853386517937764e-06,
+      "loss": 0.4518,
+      "step": 9807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1702659241151738,
+      "learning_rate": 2.8516292039416847e-06,
+      "loss": 0.4775,
+      "step": 9808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1523150424290927,
+      "learning_rate": 2.8498723412591357e-06,
+      "loss": 0.4938,
+      "step": 9809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1364860689792136,
+      "learning_rate": 2.848115930001034e-06,
+      "loss": 0.4626,
+      "step": 9810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2122917232370285,
+      "learning_rate": 2.8463599702782764e-06,
+      "loss": 0.5119,
+      "step": 9811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2063386281714108,
+      "learning_rate": 2.8446044622017223e-06,
+      "loss": 0.4895,
+      "step": 9812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.276436206262035,
+      "learning_rate": 2.842849405882202e-06,
+      "loss": 0.5402,
+      "step": 9813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0930694097448517,
+      "learning_rate": 2.841094801430524e-06,
+      "loss": 0.482,
+      "step": 9814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1627584744302293,
+      "learning_rate": 2.839340648957467e-06,
+      "loss": 0.5275,
+      "step": 9815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4099065078690134,
+      "learning_rate": 2.8375869485737782e-06,
+      "loss": 0.6152,
+      "step": 9816
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3101427617797599,
+      "learning_rate": 2.8358337003901826e-06,
+      "loss": 0.5267,
+      "step": 9817
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0559675572642508,
+      "learning_rate": 2.8340809045173646e-06,
+      "loss": 0.4663,
+      "step": 9818
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.266785266410155,
+      "learning_rate": 2.83232856106599e-06,
+      "loss": 0.5083,
+      "step": 9819
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1709059714233228,
+      "learning_rate": 2.830576670146694e-06,
+      "loss": 0.4736,
+      "step": 9820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2864326237671966,
+      "learning_rate": 2.828825231870085e-06,
+      "loss": 0.5468,
+      "step": 9821
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.312573067584006,
+      "learning_rate": 2.827074246346737e-06,
+      "loss": 0.5718,
+      "step": 9822
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.201206212987339,
+      "learning_rate": 2.825323713687197e-06,
+      "loss": 0.4914,
+      "step": 9823
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2015729033140081,
+      "learning_rate": 2.823573634001987e-06,
+      "loss": 0.4981,
+      "step": 9824
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2069773739145497,
+      "learning_rate": 2.821824007401599e-06,
+      "loss": 0.5177,
+      "step": 9825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.305339216452509,
+      "learning_rate": 2.8200748339964966e-06,
+      "loss": 0.5806,
+      "step": 9826
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1048882244072646,
+      "learning_rate": 2.818326113897115e-06,
+      "loss": 0.4816,
+      "step": 9827
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9766718688756859,
+      "learning_rate": 2.8165778472138572e-06,
+      "loss": 0.4057,
+      "step": 9828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.307712315011487,
+      "learning_rate": 2.8148300340571e-06,
+      "loss": 0.5007,
+      "step": 9829
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1727222177552468,
+      "learning_rate": 2.813082674537194e-06,
+      "loss": 0.4607,
+      "step": 9830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2639675354621294,
+      "learning_rate": 2.8113357687644615e-06,
+      "loss": 0.5698,
+      "step": 9831
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1162285590595595,
+      "learning_rate": 2.8095893168491907e-06,
+      "loss": 0.4915,
+      "step": 9832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2145241565367146,
+      "learning_rate": 2.8078433189016406e-06,
+      "loss": 0.539,
+      "step": 9833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1205390860648494,
+      "learning_rate": 2.8060977750320485e-06,
+      "loss": 0.4505,
+      "step": 9834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2136081409939452,
+      "learning_rate": 2.8043526853506187e-06,
+      "loss": 0.5186,
+      "step": 9835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2579950235750537,
+      "learning_rate": 2.80260804996753e-06,
+      "loss": 0.5054,
+      "step": 9836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0470881458621706,
+      "learning_rate": 2.8008638689929314e-06,
+      "loss": 0.4611,
+      "step": 9837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2401703585225536,
+      "learning_rate": 2.799120142536935e-06,
+      "loss": 0.5706,
+      "step": 9838
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2317386916028414,
+      "learning_rate": 2.7973768707096373e-06,
+      "loss": 0.5346,
+      "step": 9839
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0838109333019748,
+      "learning_rate": 2.795634053621098e-06,
+      "loss": 0.4016,
+      "step": 9840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.273861867018515,
+      "learning_rate": 2.793891691381353e-06,
+      "loss": 0.4628,
+      "step": 9841
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.048437347639502,
+      "learning_rate": 2.792149784100404e-06,
+      "loss": 0.4574,
+      "step": 9842
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1666399180661071,
+      "learning_rate": 2.790408331888225e-06,
+      "loss": 0.5149,
+      "step": 9843
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1567957337693864,
+      "learning_rate": 2.788667334854764e-06,
+      "loss": 0.4376,
+      "step": 9844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1276946115965643,
+      "learning_rate": 2.78692679310994e-06,
+      "loss": 0.4975,
+      "step": 9845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1262328492276015,
+      "learning_rate": 2.7851867067636407e-06,
+      "loss": 0.4986,
+      "step": 9846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.166700254819657,
+      "learning_rate": 2.783447075925729e-06,
+      "loss": 0.4907,
+      "step": 9847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2237907727845774,
+      "learning_rate": 2.78170790070604e-06,
+      "loss": 0.5076,
+      "step": 9848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2771501773601277,
+      "learning_rate": 2.779969181214368e-06,
+      "loss": 0.5389,
+      "step": 9849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1612377327885974,
+      "learning_rate": 2.7782309175604937e-06,
+      "loss": 0.4894,
+      "step": 9850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2163480417660855,
+      "learning_rate": 2.7764931098541627e-06,
+      "loss": 0.5454,
+      "step": 9851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1415034530159667,
+      "learning_rate": 2.7747557582050878e-06,
+      "loss": 0.5224,
+      "step": 9852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1984250303015094,
+      "learning_rate": 2.7730188627229617e-06,
+      "loss": 0.5258,
+      "step": 9853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2364348113927364,
+      "learning_rate": 2.7712824235174384e-06,
+      "loss": 0.5428,
+      "step": 9854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1589806248392363,
+      "learning_rate": 2.769546440698151e-06,
+      "loss": 0.4916,
+      "step": 9855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.23084618970854,
+      "learning_rate": 2.767810914374701e-06,
+      "loss": 0.5467,
+      "step": 9856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2831385115754264,
+      "learning_rate": 2.7660758446566616e-06,
+      "loss": 0.5767,
+      "step": 9857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2239159380557223,
+      "learning_rate": 2.7643412316535788e-06,
+      "loss": 0.4891,
+      "step": 9858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2424230772035525,
+      "learning_rate": 2.7626070754749623e-06,
+      "loss": 0.5698,
+      "step": 9859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2754815669445625,
+      "learning_rate": 2.7608733762303007e-06,
+      "loss": 0.5527,
+      "step": 9860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2197909189293077,
+      "learning_rate": 2.7591401340290546e-06,
+      "loss": 0.542,
+      "step": 9861
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2989702419458093,
+      "learning_rate": 2.7574073489806473e-06,
+      "loss": 0.5575,
+      "step": 9862
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1437430418042591,
+      "learning_rate": 2.7556750211944848e-06,
+      "loss": 0.5385,
+      "step": 9863
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2048709635678325,
+      "learning_rate": 2.7539431507799298e-06,
+      "loss": 0.5523,
+      "step": 9864
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2138302123007683,
+      "learning_rate": 2.752211737846329e-06,
+      "loss": 0.5013,
+      "step": 9865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1721150978811634,
+      "learning_rate": 2.7504807825029946e-06,
+      "loss": 0.5307,
+      "step": 9866
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1296348464534995,
+      "learning_rate": 2.7487502848592107e-06,
+      "loss": 0.5224,
+      "step": 9867
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0287660102015583,
+      "learning_rate": 2.7470202450242368e-06,
+      "loss": 0.4638,
+      "step": 9868
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1097480254236276,
+      "learning_rate": 2.745290663107292e-06,
+      "loss": 0.5129,
+      "step": 9869
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.185768722566527,
+      "learning_rate": 2.7435615392175763e-06,
+      "loss": 0.5038,
+      "step": 9870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1983693746329496,
+      "learning_rate": 2.741832873464262e-06,
+      "loss": 0.533,
+      "step": 9871
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2524288898162346,
+      "learning_rate": 2.7401046659564833e-06,
+      "loss": 0.5055,
+      "step": 9872
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2555332264607286,
+      "learning_rate": 2.7383769168033557e-06,
+      "loss": 0.5082,
+      "step": 9873
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2141966737029612,
+      "learning_rate": 2.736649626113955e-06,
+      "loss": 0.5693,
+      "step": 9874
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.147146108738421,
+      "learning_rate": 2.7349227939973388e-06,
+      "loss": 0.4981,
+      "step": 9875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3258379823486097,
+      "learning_rate": 2.7331964205625282e-06,
+      "loss": 0.5669,
+      "step": 9876
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2285800061845278,
+      "learning_rate": 2.7314705059185196e-06,
+      "loss": 0.4827,
+      "step": 9877
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1998316289083573,
+      "learning_rate": 2.7297450501742817e-06,
+      "loss": 0.521,
+      "step": 9878
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.197441449097756,
+      "learning_rate": 2.728020053438746e-06,
+      "loss": 0.496,
+      "step": 9879
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1359180370064426,
+      "learning_rate": 2.7262955158208215e-06,
+      "loss": 0.4892,
+      "step": 9880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1325942256709565,
+      "learning_rate": 2.724571437429393e-06,
+      "loss": 0.4609,
+      "step": 9881
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1403616509752306,
+      "learning_rate": 2.722847818373302e-06,
+      "loss": 0.5117,
+      "step": 9882
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2522275626864017,
+      "learning_rate": 2.721124658761376e-06,
+      "loss": 0.5464,
+      "step": 9883
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.191775105395397,
+      "learning_rate": 2.7194019587024024e-06,
+      "loss": 0.5021,
+      "step": 9884
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2361227289765888,
+      "learning_rate": 2.717679718305145e-06,
+      "loss": 0.537,
+      "step": 9885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2553181527665496,
+      "learning_rate": 2.7159579376783397e-06,
+      "loss": 0.5136,
+      "step": 9886
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2008053103973155,
+      "learning_rate": 2.7142366169306898e-06,
+      "loss": 0.5227,
+      "step": 9887
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2377461140509411,
+      "learning_rate": 2.712515756170876e-06,
+      "loss": 0.549,
+      "step": 9888
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.210476301412687,
+      "learning_rate": 2.710795355507537e-06,
+      "loss": 0.4939,
+      "step": 9889
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.238102319346952,
+      "learning_rate": 2.709075415049298e-06,
+      "loss": 0.5496,
+      "step": 9890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1124386995550088,
+      "learning_rate": 2.7073559349047406e-06,
+      "loss": 0.4835,
+      "step": 9891
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.169190906038654,
+      "learning_rate": 2.705636915182429e-06,
+      "loss": 0.5328,
+      "step": 9892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1445753590283219,
+      "learning_rate": 2.7039183559908954e-06,
+      "loss": 0.4852,
+      "step": 9893
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2546000715034669,
+      "learning_rate": 2.702200257438636e-06,
+      "loss": 0.4741,
+      "step": 9894
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1876169448288676,
+      "learning_rate": 2.700482619634126e-06,
+      "loss": 0.4768,
+      "step": 9895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1513088946028456,
+      "learning_rate": 2.698765442685809e-06,
+      "loss": 0.4838,
+      "step": 9896
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1666610184033224,
+      "learning_rate": 2.6970487267020985e-06,
+      "loss": 0.5128,
+      "step": 9897
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0987791786181917,
+      "learning_rate": 2.695332471791384e-06,
+      "loss": 0.5172,
+      "step": 9898
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2175867446922293,
+      "learning_rate": 2.6936166780620143e-06,
+      "loss": 0.4417,
+      "step": 9899
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1778550220646553,
+      "learning_rate": 2.691901345622322e-06,
+      "loss": 0.5309,
+      "step": 9900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2065606048418924,
+      "learning_rate": 2.6901864745806004e-06,
+      "loss": 0.533,
+      "step": 9901
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2534833058119357,
+      "learning_rate": 2.68847206504512e-06,
+      "loss": 0.5359,
+      "step": 9902
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1870311514214298,
+      "learning_rate": 2.6867581171241207e-06,
+      "loss": 0.5264,
+      "step": 9903
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2038317994621088,
+      "learning_rate": 2.685044630925816e-06,
+      "loss": 0.4904,
+      "step": 9904
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.201498492734855,
+      "learning_rate": 2.6833316065583805e-06,
+      "loss": 0.4722,
+      "step": 9905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1909770755403253,
+      "learning_rate": 2.6816190441299695e-06,
+      "loss": 0.5363,
+      "step": 9906
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3065477460800525,
+      "learning_rate": 2.6799069437487067e-06,
+      "loss": 0.5407,
+      "step": 9907
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2176764234675306,
+      "learning_rate": 2.678195305522686e-06,
+      "loss": 0.4867,
+      "step": 9908
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.254108068083646,
+      "learning_rate": 2.676484129559973e-06,
+      "loss": 0.4984,
+      "step": 9909
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2363606670128513,
+      "learning_rate": 2.6747734159686012e-06,
+      "loss": 0.5248,
+      "step": 9910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2031140512736145,
+      "learning_rate": 2.6730631648565753e-06,
+      "loss": 0.4716,
+      "step": 9911
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1020562438423822,
+      "learning_rate": 2.6713533763318724e-06,
+      "loss": 0.4827,
+      "step": 9912
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1987326068431798,
+      "learning_rate": 2.6696440505024423e-06,
+      "loss": 0.4959,
+      "step": 9913
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1936124792326168,
+      "learning_rate": 2.667935187476206e-06,
+      "loss": 0.4999,
+      "step": 9914
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2021843341639942,
+      "learning_rate": 2.666226787361046e-06,
+      "loss": 0.5125,
+      "step": 9915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2260604122330965,
+      "learning_rate": 2.6645188502648266e-06,
+      "loss": 0.5427,
+      "step": 9916
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1891810915704066,
+      "learning_rate": 2.662811376295379e-06,
+      "loss": 0.4743,
+      "step": 9917
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2013367578494354,
+      "learning_rate": 2.661104365560504e-06,
+      "loss": 0.545,
+      "step": 9918
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2662350161721243,
+      "learning_rate": 2.6593978181679758e-06,
+      "loss": 0.5615,
+      "step": 9919
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.227243319554038,
+      "learning_rate": 2.657691734225537e-06,
+      "loss": 0.5301,
+      "step": 9920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3088219287128833,
+      "learning_rate": 2.655986113840897e-06,
+      "loss": 0.5568,
+      "step": 9921
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1906285483640324,
+      "learning_rate": 2.6542809571217445e-06,
+      "loss": 0.4635,
+      "step": 9922
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2277681778742962,
+      "learning_rate": 2.6525762641757336e-06,
+      "loss": 0.5062,
+      "step": 9923
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1496360389450744,
+      "learning_rate": 2.650872035110493e-06,
+      "loss": 0.5044,
+      "step": 9924
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.084056020207639,
+      "learning_rate": 2.6491682700336165e-06,
+      "loss": 0.4536,
+      "step": 9925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1961519871243682,
+      "learning_rate": 2.6474649690526697e-06,
+      "loss": 0.5105,
+      "step": 9926
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1817719693918878,
+      "learning_rate": 2.645762132275196e-06,
+      "loss": 0.4965,
+      "step": 9927
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2424026878672974,
+      "learning_rate": 2.6440597598087005e-06,
+      "loss": 0.5254,
+      "step": 9928
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1675381584384779,
+      "learning_rate": 2.642357851760666e-06,
+      "loss": 0.4911,
+      "step": 9929
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.230200408255395,
+      "learning_rate": 2.640656408238542e-06,
+      "loss": 0.5253,
+      "step": 9930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3127762185949048,
+      "learning_rate": 2.6389554293497455e-06,
+      "loss": 0.5661,
+      "step": 9931
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.173912311905404,
+      "learning_rate": 2.6372549152016703e-06,
+      "loss": 0.4749,
+      "step": 9932
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1288718248022118,
+      "learning_rate": 2.6355548659016796e-06,
+      "loss": 0.4752,
+      "step": 9933
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2205444721119403,
+      "learning_rate": 2.633855281557108e-06,
+      "loss": 0.4786,
+      "step": 9934
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2400897565012714,
+      "learning_rate": 2.6321561622752543e-06,
+      "loss": 0.4998,
+      "step": 9935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.226849904981837,
+      "learning_rate": 2.6304575081633944e-06,
+      "loss": 0.4727,
+      "step": 9936
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2553298332113563,
+      "learning_rate": 2.628759319328774e-06,
+      "loss": 0.4812,
+      "step": 9937
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1466965747045235,
+      "learning_rate": 2.6270615958786094e-06,
+      "loss": 0.4982,
+      "step": 9938
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1735122433362208,
+      "learning_rate": 2.625364337920088e-06,
+      "loss": 0.527,
+      "step": 9939
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1700131694345968,
+      "learning_rate": 2.6236675455603634e-06,
+      "loss": 0.4744,
+      "step": 9940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1405423800710872,
+      "learning_rate": 2.6219712189065616e-06,
+      "loss": 0.4521,
+      "step": 9941
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1533850668640135,
+      "learning_rate": 2.6202753580657813e-06,
+      "loss": 0.5114,
+      "step": 9942
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2608868953945076,
+      "learning_rate": 2.6185799631450926e-06,
+      "loss": 0.534,
+      "step": 9943
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1812717839407174,
+      "learning_rate": 2.6168850342515375e-06,
+      "loss": 0.4902,
+      "step": 9944
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1256638793330935,
+      "learning_rate": 2.6151905714921187e-06,
+      "loss": 0.489,
+      "step": 9945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3486541591670242,
+      "learning_rate": 2.6134965749738195e-06,
+      "loss": 0.5749,
+      "step": 9946
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1043014564054794,
+      "learning_rate": 2.61180304480359e-06,
+      "loss": 0.4637,
+      "step": 9947
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1245948273809971,
+      "learning_rate": 2.6101099810883535e-06,
+      "loss": 0.5029,
+      "step": 9948
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3349345943416269,
+      "learning_rate": 2.6084173839350036e-06,
+      "loss": 0.5671,
+      "step": 9949
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0676599755229847,
+      "learning_rate": 2.6067252534503996e-06,
+      "loss": 0.4801,
+      "step": 9950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.263210064846937,
+      "learning_rate": 2.6050335897413713e-06,
+      "loss": 0.492,
+      "step": 9951
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2609827593654122,
+      "learning_rate": 2.6033423929147263e-06,
+      "loss": 0.5172,
+      "step": 9952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2073606270071737,
+      "learning_rate": 2.6016516630772372e-06,
+      "loss": 0.4781,
+      "step": 9953
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2026512402220375,
+      "learning_rate": 2.5999614003356523e-06,
+      "loss": 0.4969,
+      "step": 9954
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1661161190925189,
+      "learning_rate": 2.5982716047966803e-06,
+      "loss": 0.5243,
+      "step": 9955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1935415175767379,
+      "learning_rate": 2.59658227656701e-06,
+      "loss": 0.4837,
+      "step": 9956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2761928961791225,
+      "learning_rate": 2.5948934157532968e-06,
+      "loss": 0.5632,
+      "step": 9957
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.129951230996921,
+      "learning_rate": 2.5932050224621685e-06,
+      "loss": 0.5282,
+      "step": 9958
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.266337394372129,
+      "learning_rate": 2.5915170968002236e-06,
+      "loss": 0.5493,
+      "step": 9959
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.108274693240437,
+      "learning_rate": 2.589829638874026e-06,
+      "loss": 0.5348,
+      "step": 9960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1932922444293494,
+      "learning_rate": 2.5881426487901127e-06,
+      "loss": 0.4623,
+      "step": 9961
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2101212734735998,
+      "learning_rate": 2.586456126654995e-06,
+      "loss": 0.496,
+      "step": 9962
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1229493737396894,
+      "learning_rate": 2.584770072575149e-06,
+      "loss": 0.494,
+      "step": 9963
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2912410475798146,
+      "learning_rate": 2.583084486657027e-06,
+      "loss": 0.507,
+      "step": 9964
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3017216854133336,
+      "learning_rate": 2.5813993690070504e-06,
+      "loss": 0.5185,
+      "step": 9965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2180455935389933,
+      "learning_rate": 2.579714719731604e-06,
+      "loss": 0.5493,
+      "step": 9966
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.266237040281112,
+      "learning_rate": 2.5780305389370507e-06,
+      "loss": 0.5473,
+      "step": 9967
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1009691953793233,
+      "learning_rate": 2.576346826729722e-06,
+      "loss": 0.4465,
+      "step": 9968
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2663953344120586,
+      "learning_rate": 2.5746635832159216e-06,
+      "loss": 0.4871,
+      "step": 9969
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1452919056538333,
+      "learning_rate": 2.572980808501919e-06,
+      "loss": 0.4802,
+      "step": 9970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2000431867616592,
+      "learning_rate": 2.571298502693954e-06,
+      "loss": 0.5243,
+      "step": 9971
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.081226366258881,
+      "learning_rate": 2.5696166658982413e-06,
+      "loss": 0.4674,
+      "step": 9972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.180248253538023,
+      "learning_rate": 2.5679352982209637e-06,
+      "loss": 0.5176,
+      "step": 9973
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.260633822983754,
+      "learning_rate": 2.5662543997682756e-06,
+      "loss": 0.5136,
+      "step": 9974
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.171092624167524,
+      "learning_rate": 2.5645739706463037e-06,
+      "loss": 0.4875,
+      "step": 9975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1673263262976492,
+      "learning_rate": 2.5628940109611356e-06,
+      "loss": 0.4814,
+      "step": 9976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2444236349644187,
+      "learning_rate": 2.5612145208188376e-06,
+      "loss": 0.573,
+      "step": 9977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3020340210795742,
+      "learning_rate": 2.5595355003254473e-06,
+      "loss": 0.5376,
+      "step": 9978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.260800951970293,
+      "learning_rate": 2.557856949586972e-06,
+      "loss": 0.537,
+      "step": 9979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2257463692150925,
+      "learning_rate": 2.5561788687093835e-06,
+      "loss": 0.4947,
+      "step": 9980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2394408082635595,
+      "learning_rate": 2.554501257798624e-06,
+      "loss": 0.4917,
+      "step": 9981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0865589257423698,
+      "learning_rate": 2.5528241169606147e-06,
+      "loss": 0.4663,
+      "step": 9982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1127785601819211,
+      "learning_rate": 2.551147446301242e-06,
+      "loss": 0.4692,
+      "step": 9983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0347785692568185,
+      "learning_rate": 2.5494712459263615e-06,
+      "loss": 0.4479,
+      "step": 9984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.279606579281968,
+      "learning_rate": 2.547795515941803e-06,
+      "loss": 0.5007,
+      "step": 9985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2124789855551736,
+      "learning_rate": 2.5461202564533603e-06,
+      "loss": 0.5167,
+      "step": 9986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2867602046919882,
+      "learning_rate": 2.544445467566802e-06,
+      "loss": 0.5212,
+      "step": 9987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.150153040032195,
+      "learning_rate": 2.5427711493878673e-06,
+      "loss": 0.5471,
+      "step": 9988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2188503517504345,
+      "learning_rate": 2.5410973020222662e-06,
+      "loss": 0.5214,
+      "step": 9989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1628562258398445,
+      "learning_rate": 2.539423925575676e-06,
+      "loss": 0.5012,
+      "step": 9990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1414263275072591,
+      "learning_rate": 2.5377510201537427e-06,
+      "loss": 0.4883,
+      "step": 9991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2317932750547442,
+      "learning_rate": 2.5360785858620863e-06,
+      "loss": 0.5331,
+      "step": 9992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2076154856613535,
+      "learning_rate": 2.534406622806298e-06,
+      "loss": 0.504,
+      "step": 9993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.160594459259228,
+      "learning_rate": 2.532735131091937e-06,
+      "loss": 0.5296,
+      "step": 9994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2041926415269755,
+      "learning_rate": 2.531064110824536e-06,
+      "loss": 0.5302,
+      "step": 9995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1928340173768832,
+      "learning_rate": 2.52939356210959e-06,
+      "loss": 0.5035,
+      "step": 9996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2604478511683186,
+      "learning_rate": 2.527723485052571e-06,
+      "loss": 0.5003,
+      "step": 9997
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1923388223867433,
+      "learning_rate": 2.526053879758924e-06,
+      "loss": 0.4868,
+      "step": 9998
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1555894562952946,
+      "learning_rate": 2.524384746334052e-06,
+      "loss": 0.5027,
+      "step": 9999
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1767387166986665,
+      "learning_rate": 2.5227160848833443e-06,
+      "loss": 0.491,
+      "step": 10000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2164645162064749,
+      "learning_rate": 2.5210478955121444e-06,
+      "loss": 0.533,
+      "step": 10001
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2196183168072343,
+      "learning_rate": 2.5193801783257763e-06,
+      "loss": 0.5334,
+      "step": 10002
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1693086625106097,
+      "learning_rate": 2.5177129334295336e-06,
+      "loss": 0.4849,
+      "step": 10003
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2579749813927619,
+      "learning_rate": 2.5160461609286766e-06,
+      "loss": 0.4923,
+      "step": 10004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3370630530965744,
+      "learning_rate": 2.514379860928441e-06,
+      "loss": 0.552,
+      "step": 10005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1293249178887756,
+      "learning_rate": 2.5127140335340217e-06,
+      "loss": 0.4526,
+      "step": 10006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1095443918631438,
+      "learning_rate": 2.511048678850595e-06,
+      "loss": 0.4645,
+      "step": 10007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2825226859452796,
+      "learning_rate": 2.5093837969833067e-06,
+      "loss": 0.5017,
+      "step": 10008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.167332759943817,
+      "learning_rate": 2.507719388037262e-06,
+      "loss": 0.4616,
+      "step": 10009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2167484769794,
+      "learning_rate": 2.5060554521175506e-06,
+      "loss": 0.4895,
+      "step": 10010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0355720692508503,
+      "learning_rate": 2.504391989329219e-06,
+      "loss": 0.4541,
+      "step": 10011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0623148588551712,
+      "learning_rate": 2.5027289997772942e-06,
+      "loss": 0.4406,
+      "step": 10012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.178867838482162,
+      "learning_rate": 2.5010664835667677e-06,
+      "loss": 0.4825,
+      "step": 10013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1962115827109674,
+      "learning_rate": 2.499404440802604e-06,
+      "loss": 0.4951,
+      "step": 10014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1557178303645517,
+      "learning_rate": 2.4977428715897357e-06,
+      "loss": 0.4658,
+      "step": 10015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1800318840536121,
+      "learning_rate": 2.496081776033069e-06,
+      "loss": 0.5266,
+      "step": 10016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1925639047034136,
+      "learning_rate": 2.494421154237473e-06,
+      "loss": 0.4502,
+      "step": 10017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.181635126778918,
+      "learning_rate": 2.4927610063077956e-06,
+      "loss": 0.5177,
+      "step": 10018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2086557363779336,
+      "learning_rate": 2.4911013323488454e-06,
+      "loss": 0.5447,
+      "step": 10019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.148304367133798,
+      "learning_rate": 2.4894421324654084e-06,
+      "loss": 0.4846,
+      "step": 10020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.113414983915189,
+      "learning_rate": 2.487783406762242e-06,
+      "loss": 0.4754,
+      "step": 10021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2629171532093313,
+      "learning_rate": 2.4861251553440645e-06,
+      "loss": 0.6041,
+      "step": 10022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0953053451064958,
+      "learning_rate": 2.4844673783155716e-06,
+      "loss": 0.4727,
+      "step": 10023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1419672971609276,
+      "learning_rate": 2.482810075781429e-06,
+      "loss": 0.4655,
+      "step": 10024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.307920231364143,
+      "learning_rate": 2.4811532478462697e-06,
+      "loss": 0.5604,
+      "step": 10025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2963349930093868,
+      "learning_rate": 2.4794968946147012e-06,
+      "loss": 0.5272,
+      "step": 10026
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.198603966402989,
+      "learning_rate": 2.4778410161912913e-06,
+      "loss": 0.507,
+      "step": 10027
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.225843765677728,
+      "learning_rate": 2.4761856126805906e-06,
+      "loss": 0.5515,
+      "step": 10028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.119766619498295,
+      "learning_rate": 2.4745306841871063e-06,
+      "loss": 0.4824,
+      "step": 10029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.188048487030218,
+      "learning_rate": 2.4728762308153264e-06,
+      "loss": 0.522,
+      "step": 10030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.187309500823671,
+      "learning_rate": 2.4712222526697083e-06,
+      "loss": 0.5183,
+      "step": 10031
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.186278216606256,
+      "learning_rate": 2.4695687498546694e-06,
+      "loss": 0.5115,
+      "step": 10032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.214572201520226,
+      "learning_rate": 2.4679157224746076e-06,
+      "loss": 0.4824,
+      "step": 10033
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1838195383982342,
+      "learning_rate": 2.4662631706338856e-06,
+      "loss": 0.5137,
+      "step": 10034
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1206735494705995,
+      "learning_rate": 2.4646110944368393e-06,
+      "loss": 0.5,
+      "step": 10035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1044551124724868,
+      "learning_rate": 2.4629594939877754e-06,
+      "loss": 0.4861,
+      "step": 10036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.165729276681926,
+      "learning_rate": 2.461308369390961e-06,
+      "loss": 0.4922,
+      "step": 10037
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2298576654095985,
+      "learning_rate": 2.4596577207506477e-06,
+      "loss": 0.5183,
+      "step": 10038
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2426334756121429,
+      "learning_rate": 2.458007548171042e-06,
+      "loss": 0.5322,
+      "step": 10039
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2618085043903091,
+      "learning_rate": 2.4563578517563314e-06,
+      "loss": 0.5362,
+      "step": 10040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2909303945781236,
+      "learning_rate": 2.4547086316106727e-06,
+      "loss": 0.5335,
+      "step": 10041
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2597374728110449,
+      "learning_rate": 2.453059887838184e-06,
+      "loss": 0.4994,
+      "step": 10042
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2404129500499466,
+      "learning_rate": 2.451411620542962e-06,
+      "loss": 0.5025,
+      "step": 10043
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2238815067068816,
+      "learning_rate": 2.4497638298290693e-06,
+      "loss": 0.4968,
+      "step": 10044
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1828750593807622,
+      "learning_rate": 2.4481165158005395e-06,
+      "loss": 0.4855,
+      "step": 10045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.171897531928569,
+      "learning_rate": 2.4464696785613805e-06,
+      "loss": 0.4971,
+      "step": 10046
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1963336547014285,
+      "learning_rate": 2.444823318215559e-06,
+      "loss": 0.4717,
+      "step": 10047
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2407457634021282,
+      "learning_rate": 2.4431774348670236e-06,
+      "loss": 0.5609,
+      "step": 10048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1881292081594819,
+      "learning_rate": 2.441532028619682e-06,
+      "loss": 0.5203,
+      "step": 10049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2974123932914685,
+      "learning_rate": 2.4398870995774194e-06,
+      "loss": 0.5319,
+      "step": 10050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.202422642985484,
+      "learning_rate": 2.4382426478440925e-06,
+      "loss": 0.5096,
+      "step": 10051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1573668575954223,
+      "learning_rate": 2.4365986735235183e-06,
+      "loss": 0.4958,
+      "step": 10052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3252900777735628,
+      "learning_rate": 2.4349551767194914e-06,
+      "loss": 0.5182,
+      "step": 10053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2774157489235207,
+      "learning_rate": 2.433312157535774e-06,
+      "loss": 0.5626,
+      "step": 10054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1322958458639376,
+      "learning_rate": 2.431669616076101e-06,
+      "loss": 0.4618,
+      "step": 10055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.114822284836688,
+      "learning_rate": 2.430027552444174e-06,
+      "loss": 0.4799,
+      "step": 10056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2591295159239135,
+      "learning_rate": 2.4283859667436615e-06,
+      "loss": 0.5261,
+      "step": 10057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.212643608948486,
+      "learning_rate": 2.4267448590782093e-06,
+      "loss": 0.5242,
+      "step": 10058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1440619329947614,
+      "learning_rate": 2.425104229551425e-06,
+      "loss": 0.5239,
+      "step": 10059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.171990147337657,
+      "learning_rate": 2.4234640782668917e-06,
+      "loss": 0.5115,
+      "step": 10060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2037450010328585,
+      "learning_rate": 2.4218244053281636e-06,
+      "loss": 0.5091,
+      "step": 10061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1519383121338802,
+      "learning_rate": 2.420185210838756e-06,
+      "loss": 0.4638,
+      "step": 10062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1667125159291576,
+      "learning_rate": 2.418546494902163e-06,
+      "loss": 0.4872,
+      "step": 10063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3481308253502409,
+      "learning_rate": 2.416908257621845e-06,
+      "loss": 0.5743,
+      "step": 10064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1603950229089637,
+      "learning_rate": 2.415270499101232e-06,
+      "loss": 0.521,
+      "step": 10065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2866470826508272,
+      "learning_rate": 2.413633219443725e-06,
+      "loss": 0.5425,
+      "step": 10066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1823853218974247,
+      "learning_rate": 2.411996418752696e-06,
+      "loss": 0.5001,
+      "step": 10067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1507584869350262,
+      "learning_rate": 2.410360097131482e-06,
+      "loss": 0.4919,
+      "step": 10068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1746519628427923,
+      "learning_rate": 2.4087242546833887e-06,
+      "loss": 0.5151,
+      "step": 10069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1421706818100408,
+      "learning_rate": 2.407088891511701e-06,
+      "loss": 0.4504,
+      "step": 10070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.199424750810467,
+      "learning_rate": 2.4054540077196644e-06,
+      "loss": 0.4759,
+      "step": 10071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1506963818987264,
+      "learning_rate": 2.403819603410502e-06,
+      "loss": 0.4707,
+      "step": 10072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2772116869863952,
+      "learning_rate": 2.4021856786873964e-06,
+      "loss": 0.5053,
+      "step": 10073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2042466421969442,
+      "learning_rate": 2.400552233653508e-06,
+      "loss": 0.4835,
+      "step": 10074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2620288007324452,
+      "learning_rate": 2.398919268411967e-06,
+      "loss": 0.53,
+      "step": 10075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.144201758546483,
+      "learning_rate": 2.3972867830658665e-06,
+      "loss": 0.4688,
+      "step": 10076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0487533912103724,
+      "learning_rate": 2.3956547777182805e-06,
+      "loss": 0.4783,
+      "step": 10077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2807614860578087,
+      "learning_rate": 2.394023252472242e-06,
+      "loss": 0.5428,
+      "step": 10078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1751886439566128,
+      "learning_rate": 2.392392207430754e-06,
+      "loss": 0.4952,
+      "step": 10079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2347904423672806,
+      "learning_rate": 2.390761642696795e-06,
+      "loss": 0.5202,
+      "step": 10080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1340833704171946,
+      "learning_rate": 2.3891315583733133e-06,
+      "loss": 0.5496,
+      "step": 10081
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.347717151095056,
+      "learning_rate": 2.387501954563225e-06,
+      "loss": 0.5304,
+      "step": 10082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.235878957719505,
+      "learning_rate": 2.385872831369411e-06,
+      "loss": 0.4912,
+      "step": 10083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2240811658326685,
+      "learning_rate": 2.38424418889473e-06,
+      "loss": 0.4682,
+      "step": 10084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2674457968835273,
+      "learning_rate": 2.382616027242005e-06,
+      "loss": 0.5136,
+      "step": 10085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2216633430442685,
+      "learning_rate": 2.3809883465140304e-06,
+      "loss": 0.461,
+      "step": 10086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2392178913056242,
+      "learning_rate": 2.3793611468135734e-06,
+      "loss": 0.5249,
+      "step": 10087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1419433395405967,
+      "learning_rate": 2.3777344282433645e-06,
+      "loss": 0.4885,
+      "step": 10088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1360705123327532,
+      "learning_rate": 2.3761081909061036e-06,
+      "loss": 0.4726,
+      "step": 10089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2173158080965867,
+      "learning_rate": 2.374482434904467e-06,
+      "loss": 0.5024,
+      "step": 10090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2974416115472984,
+      "learning_rate": 2.372857160341098e-06,
+      "loss": 0.4977,
+      "step": 10091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.209564217838977,
+      "learning_rate": 2.371232367318609e-06,
+      "loss": 0.5027,
+      "step": 10092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1404234303078846,
+      "learning_rate": 2.369608055939576e-06,
+      "loss": 0.464,
+      "step": 10093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2176849895988486,
+      "learning_rate": 2.3679842263065554e-06,
+      "loss": 0.5141,
+      "step": 10094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1304284099770816,
+      "learning_rate": 2.366360878522067e-06,
+      "loss": 0.5312,
+      "step": 10095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2151739524208165,
+      "learning_rate": 2.3647380126886033e-06,
+      "loss": 0.4729,
+      "step": 10096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0972838397906044,
+      "learning_rate": 2.363115628908619e-06,
+      "loss": 0.4701,
+      "step": 10097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2570521739975058,
+      "learning_rate": 2.3614937272845484e-06,
+      "loss": 0.5217,
+      "step": 10098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1563948463235374,
+      "learning_rate": 2.3598723079187848e-06,
+      "loss": 0.4791,
+      "step": 10099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.211298685660256,
+      "learning_rate": 2.358251370913701e-06,
+      "loss": 0.4732,
+      "step": 10100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0808573941680326,
+      "learning_rate": 2.356630916371635e-06,
+      "loss": 0.5037,
+      "step": 10101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2364363540114214,
+      "learning_rate": 2.3550109443948967e-06,
+      "loss": 0.501,
+      "step": 10102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0159774755511717,
+      "learning_rate": 2.353391455085756e-06,
+      "loss": 0.4007,
+      "step": 10103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1697519016499072,
+      "learning_rate": 2.3517724485464655e-06,
+      "loss": 0.4867,
+      "step": 10104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.248692639458742,
+      "learning_rate": 2.3501539248792406e-06,
+      "loss": 0.502,
+      "step": 10105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1643580119213923,
+      "learning_rate": 2.348535884186267e-06,
+      "loss": 0.4728,
+      "step": 10106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.188584585303675,
+      "learning_rate": 2.3469183265696984e-06,
+      "loss": 0.4422,
+      "step": 10107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2102148545458833,
+      "learning_rate": 2.3453012521316633e-06,
+      "loss": 0.4905,
+      "step": 10108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.209539184465441,
+      "learning_rate": 2.343684660974249e-06,
+      "loss": 0.5305,
+      "step": 10109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.075547828366324,
+      "learning_rate": 2.3420685531995247e-06,
+      "loss": 0.4604,
+      "step": 10110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1997112045867584,
+      "learning_rate": 2.340452928909522e-06,
+      "loss": 0.5389,
+      "step": 10111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2543143205549179,
+      "learning_rate": 2.3388377882062472e-06,
+      "loss": 0.5065,
+      "step": 10112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1929581337526975,
+      "learning_rate": 2.337223131191666e-06,
+      "loss": 0.4766,
+      "step": 10113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1575132634533807,
+      "learning_rate": 2.335608957967723e-06,
+      "loss": 0.4926,
+      "step": 10114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3458825422252627,
+      "learning_rate": 2.33399526863633e-06,
+      "loss": 0.5577,
+      "step": 10115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1670118513804872,
+      "learning_rate": 2.33238206329937e-06,
+      "loss": 0.4899,
+      "step": 10116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1514919947827544,
+      "learning_rate": 2.3307693420586873e-06,
+      "loss": 0.4623,
+      "step": 10117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1456185341843501,
+      "learning_rate": 2.3291571050161066e-06,
+      "loss": 0.4777,
+      "step": 10118
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1668270999862007,
+      "learning_rate": 2.3275453522734116e-06,
+      "loss": 0.5206,
+      "step": 10119
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2782400643708445,
+      "learning_rate": 2.325934083932362e-06,
+      "loss": 0.5409,
+      "step": 10120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.11572265625,
+      "learning_rate": 2.3243233000946874e-06,
+      "loss": 0.4567,
+      "step": 10121
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1744299946688732,
+      "learning_rate": 2.3227130008620847e-06,
+      "loss": 0.4407,
+      "step": 10122
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2944843824762486,
+      "learning_rate": 2.321103186336222e-06,
+      "loss": 0.5814,
+      "step": 10123
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3216213984071001,
+      "learning_rate": 2.319493856618731e-06,
+      "loss": 0.5074,
+      "step": 10124
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1083881663380104,
+      "learning_rate": 2.3178850118112185e-06,
+      "loss": 0.4991,
+      "step": 10125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3344712269590773,
+      "learning_rate": 2.3162766520152624e-06,
+      "loss": 0.5435,
+      "step": 10126
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.136803482935145,
+      "learning_rate": 2.3146687773324017e-06,
+      "loss": 0.4609,
+      "step": 10127
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1339425074318483,
+      "learning_rate": 2.313061387864155e-06,
+      "loss": 0.4937,
+      "step": 10128
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2276166528717671,
+      "learning_rate": 2.311454483711999e-06,
+      "loss": 0.466,
+      "step": 10129
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2216988614633686,
+      "learning_rate": 2.309848064977389e-06,
+      "loss": 0.5624,
+      "step": 10130
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1410175327388248,
+      "learning_rate": 2.3082421317617463e-06,
+      "loss": 0.5206,
+      "step": 10131
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1642994991442626,
+      "learning_rate": 2.3066366841664633e-06,
+      "loss": 0.5073,
+      "step": 10132
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2936334658858177,
+      "learning_rate": 2.3050317222929006e-06,
+      "loss": 0.4881,
+      "step": 10133
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.218495611393429,
+      "learning_rate": 2.3034272462423846e-06,
+      "loss": 0.4935,
+      "step": 10134
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.191787958752573,
+      "learning_rate": 2.3018232561162144e-06,
+      "loss": 0.5166,
+      "step": 10135
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2516170532681377,
+      "learning_rate": 2.3002197520156634e-06,
+      "loss": 0.5487,
+      "step": 10136
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0504019240469935,
+      "learning_rate": 2.2986167340419606e-06,
+      "loss": 0.472,
+      "step": 10137
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1487923936627922,
+      "learning_rate": 2.2970142022963216e-06,
+      "loss": 0.4904,
+      "step": 10138
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2419663718273535,
+      "learning_rate": 2.295412156879915e-06,
+      "loss": 0.5323,
+      "step": 10139
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2529952403021525,
+      "learning_rate": 2.2938105978938897e-06,
+      "loss": 0.5181,
+      "step": 10140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1726901461831587,
+      "learning_rate": 2.2922095254393594e-06,
+      "loss": 0.4811,
+      "step": 10141
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2230176787545148,
+      "learning_rate": 2.290608939617408e-06,
+      "loss": 0.5283,
+      "step": 10142
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2528957204756141,
+      "learning_rate": 2.2890088405290935e-06,
+      "loss": 0.4803,
+      "step": 10143
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1980383216893382,
+      "learning_rate": 2.2874092282754313e-06,
+      "loss": 0.5198,
+      "step": 10144
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1562906722055037,
+      "learning_rate": 2.2858101029574164e-06,
+      "loss": 0.4644,
+      "step": 10145
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1054367761739514,
+      "learning_rate": 2.284211464676013e-06,
+      "loss": 0.4602,
+      "step": 10146
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1415859511770745,
+      "learning_rate": 2.2826133135321437e-06,
+      "loss": 0.4697,
+      "step": 10147
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2473425750937144,
+      "learning_rate": 2.2810156496267165e-06,
+      "loss": 0.5214,
+      "step": 10148
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2849709977316985,
+      "learning_rate": 2.2794184730605926e-06,
+      "loss": 0.5002,
+      "step": 10149
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.230826867684282,
+      "learning_rate": 2.277821783934614e-06,
+      "loss": 0.5676,
+      "step": 10150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1926225300983107,
+      "learning_rate": 2.276225582349587e-06,
+      "loss": 0.4976,
+      "step": 10151
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2403045874718333,
+      "learning_rate": 2.2746298684062884e-06,
+      "loss": 0.5158,
+      "step": 10152
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2616096662010758,
+      "learning_rate": 2.273034642205467e-06,
+      "loss": 0.5316,
+      "step": 10153
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1504857488658893,
+      "learning_rate": 2.2714399038478317e-06,
+      "loss": 0.4866,
+      "step": 10154
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1106666245439514,
+      "learning_rate": 2.2698456534340694e-06,
+      "loss": 0.4931,
+      "step": 10155
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2648908110375974,
+      "learning_rate": 2.2682518910648353e-06,
+      "loss": 0.5406,
+      "step": 10156
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0745793517621873,
+      "learning_rate": 2.266658616840748e-06,
+      "loss": 0.4133,
+      "step": 10157
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.27854405740696,
+      "learning_rate": 2.265065830862404e-06,
+      "loss": 0.5401,
+      "step": 10158
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3219401692948791,
+      "learning_rate": 2.2634735332303583e-06,
+      "loss": 0.5505,
+      "step": 10159
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2212095141069266,
+      "learning_rate": 2.261881724045143e-06,
+      "loss": 0.4825,
+      "step": 10160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.091805118536494,
+      "learning_rate": 2.2602904034072583e-06,
+      "loss": 0.4823,
+      "step": 10161
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2066577716274733,
+      "learning_rate": 2.2586995714171712e-06,
+      "loss": 0.5196,
+      "step": 10162
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2136639327423837,
+      "learning_rate": 2.257109228175324e-06,
+      "loss": 0.5029,
+      "step": 10163
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1878321333750181,
+      "learning_rate": 2.2555193737821156e-06,
+      "loss": 0.5324,
+      "step": 10164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3371091021872668,
+      "learning_rate": 2.253930008337927e-06,
+      "loss": 0.5146,
+      "step": 10165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2784012549574402,
+      "learning_rate": 2.252341131943102e-06,
+      "loss": 0.5207,
+      "step": 10166
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1155664914814352,
+      "learning_rate": 2.250752744697953e-06,
+      "loss": 0.541,
+      "step": 10167
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.193134892354234,
+      "learning_rate": 2.249164846702766e-06,
+      "loss": 0.4344,
+      "step": 10168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.227594852317764,
+      "learning_rate": 2.247577438057789e-06,
+      "loss": 0.5493,
+      "step": 10169
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2751448249842616,
+      "learning_rate": 2.2459905188632446e-06,
+      "loss": 0.5256,
+      "step": 10170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2167122261888477,
+      "learning_rate": 2.244404089219325e-06,
+      "loss": 0.5444,
+      "step": 10171
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2273949392074224,
+      "learning_rate": 2.242818149226189e-06,
+      "loss": 0.5473,
+      "step": 10172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2131404895766427,
+      "learning_rate": 2.2412326989839673e-06,
+      "loss": 0.489,
+      "step": 10173
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.165382046986016,
+      "learning_rate": 2.2396477385927527e-06,
+      "loss": 0.4718,
+      "step": 10174
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4393539083958824,
+      "learning_rate": 2.238063268152615e-06,
+      "loss": 0.5621,
+      "step": 10175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1952756489577088,
+      "learning_rate": 2.236479287763591e-06,
+      "loss": 0.5411,
+      "step": 10176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2702662761015613,
+      "learning_rate": 2.2348957975256825e-06,
+      "loss": 0.52,
+      "step": 10177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2513012788461038,
+      "learning_rate": 2.2333127975388646e-06,
+      "loss": 0.5292,
+      "step": 10178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2521675866459296,
+      "learning_rate": 2.2317302879030833e-06,
+      "loss": 0.5621,
+      "step": 10179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.13786158840586,
+      "learning_rate": 2.230148268718245e-06,
+      "loss": 0.4824,
+      "step": 10180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.173078046620181,
+      "learning_rate": 2.228566740084234e-06,
+      "loss": 0.4786,
+      "step": 10181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1558363406511902,
+      "learning_rate": 2.2269857021009e-06,
+      "loss": 0.5146,
+      "step": 10182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1292032029456813,
+      "learning_rate": 2.225405154868062e-06,
+      "loss": 0.5,
+      "step": 10183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1289229342273814,
+      "learning_rate": 2.2238250984855112e-06,
+      "loss": 0.4808,
+      "step": 10184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2538177839308264,
+      "learning_rate": 2.2222455330529993e-06,
+      "loss": 0.5224,
+      "step": 10185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3138769737753615,
+      "learning_rate": 2.2206664586702566e-06,
+      "loss": 0.5375,
+      "step": 10186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1931142602100457,
+      "learning_rate": 2.219087875436974e-06,
+      "loss": 0.4765,
+      "step": 10187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.145986211577279,
+      "learning_rate": 2.2175097834528183e-06,
+      "loss": 0.4783,
+      "step": 10188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.100529883378034,
+      "learning_rate": 2.2159321828174252e-06,
+      "loss": 0.4952,
+      "step": 10189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1984984878479574,
+      "learning_rate": 2.214355073630391e-06,
+      "loss": 0.464,
+      "step": 10190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.195610408185424,
+      "learning_rate": 2.212778455991289e-06,
+      "loss": 0.4916,
+      "step": 10191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2169761464422881,
+      "learning_rate": 2.211202329999661e-06,
+      "loss": 0.4983,
+      "step": 10192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2559834800137961,
+      "learning_rate": 2.2096266957550138e-06,
+      "loss": 0.5574,
+      "step": 10193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1271198750963534,
+      "learning_rate": 2.208051553356829e-06,
+      "loss": 0.4698,
+      "step": 10194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1662122091076532,
+      "learning_rate": 2.2064769029045497e-06,
+      "loss": 0.497,
+      "step": 10195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2296533698613163,
+      "learning_rate": 2.2049027444975934e-06,
+      "loss": 0.5144,
+      "step": 10196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2364310030446166,
+      "learning_rate": 2.2033290782353434e-06,
+      "loss": 0.5392,
+      "step": 10197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1779016277935124,
+      "learning_rate": 2.2017559042171534e-06,
+      "loss": 0.5042,
+      "step": 10198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2021413968683448,
+      "learning_rate": 2.2001832225423493e-06,
+      "loss": 0.5654,
+      "step": 10199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1834056440386747,
+      "learning_rate": 2.1986110333102175e-06,
+      "loss": 0.4727,
+      "step": 10200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.331240198161974,
+      "learning_rate": 2.1970393366200216e-06,
+      "loss": 0.479,
+      "step": 10201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1756101971778818,
+      "learning_rate": 2.195468132570989e-06,
+      "loss": 0.4601,
+      "step": 10202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0827746540187844,
+      "learning_rate": 2.193897421262321e-06,
+      "loss": 0.4897,
+      "step": 10203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1487147716645416,
+      "learning_rate": 2.1923272027931853e-06,
+      "loss": 0.4507,
+      "step": 10204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2662310620894188,
+      "learning_rate": 2.1907574772627116e-06,
+      "loss": 0.531,
+      "step": 10205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2257299331027671,
+      "learning_rate": 2.189188244770013e-06,
+      "loss": 0.4869,
+      "step": 10206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1055489770468852,
+      "learning_rate": 2.187619505414156e-06,
+      "loss": 0.5016,
+      "step": 10207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2545303265283958,
+      "learning_rate": 2.186051259294185e-06,
+      "loss": 0.52,
+      "step": 10208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1895532424960171,
+      "learning_rate": 2.1844835065091165e-06,
+      "loss": 0.5058,
+      "step": 10209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.190883634415636,
+      "learning_rate": 2.1829162471579234e-06,
+      "loss": 0.5163,
+      "step": 10210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0893931123734641,
+      "learning_rate": 2.18134948133956e-06,
+      "loss": 0.4837,
+      "step": 10211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1292518166017984,
+      "learning_rate": 2.1797832091529414e-06,
+      "loss": 0.5445,
+      "step": 10212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1243821142712682,
+      "learning_rate": 2.178217430696956e-06,
+      "loss": 0.5161,
+      "step": 10213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0830331838953589,
+      "learning_rate": 2.1766521460704627e-06,
+      "loss": 0.4431,
+      "step": 10214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2119092918136107,
+      "learning_rate": 2.1750873553722796e-06,
+      "loss": 0.4775,
+      "step": 10215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0401988305247927,
+      "learning_rate": 2.1735230587012057e-06,
+      "loss": 0.4159,
+      "step": 10216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.318567737079112,
+      "learning_rate": 2.1719592561559978e-06,
+      "loss": 0.5214,
+      "step": 10217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0961440995847413,
+      "learning_rate": 2.1703959478353886e-06,
+      "loss": 0.4706,
+      "step": 10218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2243958793600707,
+      "learning_rate": 2.168833133838082e-06,
+      "loss": 0.4916,
+      "step": 10219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2505341818956475,
+      "learning_rate": 2.167270814262741e-06,
+      "loss": 0.4605,
+      "step": 10220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3083759639055283,
+      "learning_rate": 2.165708989208004e-06,
+      "loss": 0.4989,
+      "step": 10221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.321453932846483,
+      "learning_rate": 2.1641476587724784e-06,
+      "loss": 0.5485,
+      "step": 10222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3265343958525244,
+      "learning_rate": 2.16258682305474e-06,
+      "loss": 0.5625,
+      "step": 10223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1665851984643831,
+      "learning_rate": 2.1610264821533323e-06,
+      "loss": 0.5387,
+      "step": 10224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1498525296899968,
+      "learning_rate": 2.159466636166765e-06,
+      "loss": 0.508,
+      "step": 10225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2169015511788683,
+      "learning_rate": 2.1579072851935222e-06,
+      "loss": 0.5239,
+      "step": 10226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3006873111197723,
+      "learning_rate": 2.1563484293320503e-06,
+      "loss": 0.5655,
+      "step": 10227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2583264077019156,
+      "learning_rate": 2.154790068680771e-06,
+      "loss": 0.4888,
+      "step": 10228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3641088209119965,
+      "learning_rate": 2.1532322033380725e-06,
+      "loss": 0.5536,
+      "step": 10229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2896213591623045,
+      "learning_rate": 2.1516748334023065e-06,
+      "loss": 0.5742,
+      "step": 10230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.266656996455841,
+      "learning_rate": 2.1501179589717993e-06,
+      "loss": 0.558,
+      "step": 10231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.23049209814778,
+      "learning_rate": 2.148561580144847e-06,
+      "loss": 0.5469,
+      "step": 10232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.083509461439479,
+      "learning_rate": 2.1470056970197085e-06,
+      "loss": 0.489,
+      "step": 10233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1973030619656757,
+      "learning_rate": 2.14545030969462e-06,
+      "loss": 0.5326,
+      "step": 10234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1525056369668465,
+      "learning_rate": 2.143895418267775e-06,
+      "loss": 0.532,
+      "step": 10235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2896594890402306,
+      "learning_rate": 2.1423410228373477e-06,
+      "loss": 0.5312,
+      "step": 10236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1524543838410315,
+      "learning_rate": 2.1407871235014675e-06,
+      "loss": 0.4639,
+      "step": 10237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3704034094412325,
+      "learning_rate": 2.139233720358246e-06,
+      "loss": 0.5722,
+      "step": 10238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2197110714932773,
+      "learning_rate": 2.137680813505756e-06,
+      "loss": 0.5238,
+      "step": 10239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.23459961211722,
+      "learning_rate": 2.1361284030420416e-06,
+      "loss": 0.5142,
+      "step": 10240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1625394650395113,
+      "learning_rate": 2.1345764890651123e-06,
+      "loss": 0.4782,
+      "step": 10241
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.232935440257451,
+      "learning_rate": 2.1330250716729484e-06,
+      "loss": 0.4937,
+      "step": 10242
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0461478483585889,
+      "learning_rate": 2.1314741509635007e-06,
+      "loss": 0.4287,
+      "step": 10243
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.117240317636678,
+      "learning_rate": 2.1299237270346885e-06,
+      "loss": 0.4419,
+      "step": 10244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0641729983681847,
+      "learning_rate": 2.1283737999843922e-06,
+      "loss": 0.4996,
+      "step": 10245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2339668081880149,
+      "learning_rate": 2.126824369910474e-06,
+      "loss": 0.5177,
+      "step": 10246
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0542395311486665,
+      "learning_rate": 2.125275436910751e-06,
+      "loss": 0.4682,
+      "step": 10247
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.286338332083872,
+      "learning_rate": 2.123727001083017e-06,
+      "loss": 0.5072,
+      "step": 10248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1217494203631044,
+      "learning_rate": 2.1221790625250336e-06,
+      "loss": 0.477,
+      "step": 10249
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1588988506260418,
+      "learning_rate": 2.1206316213345334e-06,
+      "loss": 0.5209,
+      "step": 10250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3158661048308031,
+      "learning_rate": 2.119084677609209e-06,
+      "loss": 0.4853,
+      "step": 10251
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2928356090579884,
+      "learning_rate": 2.1175382314467285e-06,
+      "loss": 0.4804,
+      "step": 10252
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.230745653754229,
+      "learning_rate": 2.1159922829447267e-06,
+      "loss": 0.4912,
+      "step": 10253
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.186337956503034,
+      "learning_rate": 2.1144468322008125e-06,
+      "loss": 0.5167,
+      "step": 10254
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1992578655412016,
+      "learning_rate": 2.11290187931255e-06,
+      "loss": 0.5538,
+      "step": 10255
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1897909752606575,
+      "learning_rate": 2.111357424377487e-06,
+      "loss": 0.5504,
+      "step": 10256
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3010672406670374,
+      "learning_rate": 2.109813467493127e-06,
+      "loss": 0.5576,
+      "step": 10257
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2899843560977982,
+      "learning_rate": 2.1082700087569517e-06,
+      "loss": 0.585,
+      "step": 10258
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2554375160951472,
+      "learning_rate": 2.106727048266406e-06,
+      "loss": 0.5399,
+      "step": 10259
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2978699326201804,
+      "learning_rate": 2.105184586118908e-06,
+      "loss": 0.5437,
+      "step": 10260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1338721745114873,
+      "learning_rate": 2.1036426224118366e-06,
+      "loss": 0.4785,
+      "step": 10261
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2578823473000358,
+      "learning_rate": 2.1021011572425466e-06,
+      "loss": 0.4409,
+      "step": 10262
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1481951503265189,
+      "learning_rate": 2.1005601907083585e-06,
+      "loss": 0.4557,
+      "step": 10263
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1707571674610573,
+      "learning_rate": 2.0990197229065636e-06,
+      "loss": 0.4925,
+      "step": 10264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3843083453869778,
+      "learning_rate": 2.097479753934415e-06,
+      "loss": 0.526,
+      "step": 10265
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0259899894493696,
+      "learning_rate": 2.095940283889143e-06,
+      "loss": 0.4505,
+      "step": 10266
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3241290846993647,
+      "learning_rate": 2.0944013128679385e-06,
+      "loss": 0.5724,
+      "step": 10267
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1714965209127777,
+      "learning_rate": 2.092862840967966e-06,
+      "loss": 0.5172,
+      "step": 10268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.181992508378791,
+      "learning_rate": 2.0913248682863583e-06,
+      "loss": 0.4796,
+      "step": 10269
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1627365343375886,
+      "learning_rate": 2.0897873949202175e-06,
+      "loss": 0.53,
+      "step": 10270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1728276259038721,
+      "learning_rate": 2.088250420966608e-06,
+      "loss": 0.4762,
+      "step": 10271
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.196731224089679,
+      "learning_rate": 2.086713946522567e-06,
+      "loss": 0.5219,
+      "step": 10272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1620583402823577,
+      "learning_rate": 2.085177971685103e-06,
+      "loss": 0.5089,
+      "step": 10273
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1714205560094304,
+      "learning_rate": 2.0836424965511913e-06,
+      "loss": 0.4839,
+      "step": 10274
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1311805282252312,
+      "learning_rate": 2.082107521217769e-06,
+      "loss": 0.456,
+      "step": 10275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.253755221674534,
+      "learning_rate": 2.080573045781753e-06,
+      "loss": 0.4884,
+      "step": 10276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0926340677330009,
+      "learning_rate": 2.0790390703400164e-06,
+      "loss": 0.4614,
+      "step": 10277
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.240784050189275,
+      "learning_rate": 2.0775055949894096e-06,
+      "loss": 0.5123,
+      "step": 10278
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2754458171591807,
+      "learning_rate": 2.0759726198267495e-06,
+      "loss": 0.535,
+      "step": 10279
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2359064959122852,
+      "learning_rate": 2.0744401449488238e-06,
+      "loss": 0.5527,
+      "step": 10280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.095415482189708,
+      "learning_rate": 2.072908170452379e-06,
+      "loss": 0.4655,
+      "step": 10281
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.235275278261656,
+      "learning_rate": 2.07137669643414e-06,
+      "loss": 0.5061,
+      "step": 10282
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1767587748796635,
+      "learning_rate": 2.0698457229907966e-06,
+      "loss": 0.4943,
+      "step": 10283
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1228389647051331,
+      "learning_rate": 2.0683152502190095e-06,
+      "loss": 0.5053,
+      "step": 10284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1843336958715984,
+      "learning_rate": 2.0667852782153996e-06,
+      "loss": 0.4801,
+      "step": 10285
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2250850064113754,
+      "learning_rate": 2.0652558070765682e-06,
+      "loss": 0.5531,
+      "step": 10286
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1554970738445292,
+      "learning_rate": 2.0637268368990727e-06,
+      "loss": 0.4522,
+      "step": 10287
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1650304674085714,
+      "learning_rate": 2.0621983677794486e-06,
+      "loss": 0.5014,
+      "step": 10288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2780542293439354,
+      "learning_rate": 2.0606703998141942e-06,
+      "loss": 0.501,
+      "step": 10289
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1134729588397931,
+      "learning_rate": 2.0591429330997793e-06,
+      "loss": 0.465,
+      "step": 10290
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.095193836352529,
+      "learning_rate": 2.0576159677326437e-06,
+      "loss": 0.468,
+      "step": 10291
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2125889008169002,
+      "learning_rate": 2.0560895038091865e-06,
+      "loss": 0.4905,
+      "step": 10292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2807207177333237,
+      "learning_rate": 2.0545635414257835e-06,
+      "loss": 0.5412,
+      "step": 10293
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1189182986039712,
+      "learning_rate": 2.053038080678781e-06,
+      "loss": 0.4888,
+      "step": 10294
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2092515098746894,
+      "learning_rate": 2.051513121664481e-06,
+      "loss": 0.5595,
+      "step": 10295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3039500727983295,
+      "learning_rate": 2.04998866447917e-06,
+      "loss": 0.5679,
+      "step": 10296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.185788276139838,
+      "learning_rate": 2.048464709219089e-06,
+      "loss": 0.5664,
+      "step": 10297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1052751674931791,
+      "learning_rate": 2.046941255980456e-06,
+      "loss": 0.5147,
+      "step": 10298
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.33807064432412,
+      "learning_rate": 2.0454183048594524e-06,
+      "loss": 0.5219,
+      "step": 10299
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2498117782026792,
+      "learning_rate": 2.0438958559522314e-06,
+      "loss": 0.5476,
+      "step": 10300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2121789290854608,
+      "learning_rate": 2.042373909354917e-06,
+      "loss": 0.4863,
+      "step": 10301
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2257827418399094,
+      "learning_rate": 2.040852465163591e-06,
+      "loss": 0.4721,
+      "step": 10302
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2895661728332928,
+      "learning_rate": 2.0393315234743116e-06,
+      "loss": 0.4944,
+      "step": 10303
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.182525152378456,
+      "learning_rate": 2.0378110843831077e-06,
+      "loss": 0.53,
+      "step": 10304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2645572349624634,
+      "learning_rate": 2.036291147985967e-06,
+      "loss": 0.5323,
+      "step": 10305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1360488962913127,
+      "learning_rate": 2.034771714378857e-06,
+      "loss": 0.4909,
+      "step": 10306
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1664564363346435,
+      "learning_rate": 2.0332527836577e-06,
+      "loss": 0.5156,
+      "step": 10307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1656944561021052,
+      "learning_rate": 2.031734355918399e-06,
+      "loss": 0.5147,
+      "step": 10308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2040978498935313,
+      "learning_rate": 2.0302164312568175e-06,
+      "loss": 0.4824,
+      "step": 10309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0852949587972098,
+      "learning_rate": 2.028699009768792e-06,
+      "loss": 0.4967,
+      "step": 10310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.133489367084751,
+      "learning_rate": 2.0271820915501273e-06,
+      "loss": 0.4768,
+      "step": 10311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2591906751055966,
+      "learning_rate": 2.025665676696589e-06,
+      "loss": 0.5354,
+      "step": 10312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2247059371818034,
+      "learning_rate": 2.0241497653039178e-06,
+      "loss": 0.5186,
+      "step": 10313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.214661710292141,
+      "learning_rate": 2.0226343574678255e-06,
+      "loss": 0.514,
+      "step": 10314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1841975519518038,
+      "learning_rate": 2.0211194532839807e-06,
+      "loss": 0.5078,
+      "step": 10315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1738868736760906,
+      "learning_rate": 2.019605052848034e-06,
+      "loss": 0.4821,
+      "step": 10316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1561511744974924,
+      "learning_rate": 2.0180911562555904e-06,
+      "loss": 0.5153,
+      "step": 10317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.144552809997178,
+      "learning_rate": 2.016577763602233e-06,
+      "loss": 0.4528,
+      "step": 10318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1583653765758537,
+      "learning_rate": 2.015064874983511e-06,
+      "loss": 0.5034,
+      "step": 10319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1670531189146744,
+      "learning_rate": 2.01355249049494e-06,
+      "loss": 0.4809,
+      "step": 10320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.149081562082276,
+      "learning_rate": 2.012040610232008e-06,
+      "loss": 0.4941,
+      "step": 10321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2771470504638869,
+      "learning_rate": 2.0105292342901617e-06,
+      "loss": 0.5262,
+      "step": 10322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1730759125785757,
+      "learning_rate": 2.009018362764825e-06,
+      "loss": 0.5145,
+      "step": 10323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2472225326324355,
+      "learning_rate": 2.007507995751391e-06,
+      "loss": 0.533,
+      "step": 10324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3087537069914892,
+      "learning_rate": 2.005998133345208e-06,
+      "loss": 0.531,
+      "step": 10325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1604079156587876,
+      "learning_rate": 2.004488775641611e-06,
+      "loss": 0.5015,
+      "step": 10326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.405106779161079,
+      "learning_rate": 2.002979922735886e-06,
+      "loss": 0.5834,
+      "step": 10327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2972890353825752,
+      "learning_rate": 2.001471574723298e-06,
+      "loss": 0.549,
+      "step": 10328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1045161479910992,
+      "learning_rate": 1.999963731699076e-06,
+      "loss": 0.4414,
+      "step": 10329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1388594144144728,
+      "learning_rate": 1.9984563937584177e-06,
+      "loss": 0.4868,
+      "step": 10330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1982704414201537,
+      "learning_rate": 1.996949560996494e-06,
+      "loss": 0.5345,
+      "step": 10331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3591611408953175,
+      "learning_rate": 1.9954432335084307e-06,
+      "loss": 0.5535,
+      "step": 10332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1413411021978321,
+      "learning_rate": 1.9939374113893353e-06,
+      "loss": 0.4713,
+      "step": 10333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2577695661526864,
+      "learning_rate": 1.992432094734279e-06,
+      "loss": 0.5294,
+      "step": 10334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.152397645886078,
+      "learning_rate": 1.9909272836382955e-06,
+      "loss": 0.4752,
+      "step": 10335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2084929865016871,
+      "learning_rate": 1.9894229781963957e-06,
+      "loss": 0.5307,
+      "step": 10336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.184222466704197,
+      "learning_rate": 1.9879191785035513e-06,
+      "loss": 0.5068,
+      "step": 10337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.288310299667858,
+      "learning_rate": 1.9864158846547054e-06,
+      "loss": 0.5117,
+      "step": 10338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1521021185522449,
+      "learning_rate": 1.9849130967447693e-06,
+      "loss": 0.4821,
+      "step": 10339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3211294502113327,
+      "learning_rate": 1.9834108148686225e-06,
+      "loss": 0.5224,
+      "step": 10340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.126520877233239,
+      "learning_rate": 1.9819090391211104e-06,
+      "loss": 0.498,
+      "step": 10341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2187549884400615,
+      "learning_rate": 1.9804077695970513e-06,
+      "loss": 0.5028,
+      "step": 10342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3013953056887446,
+      "learning_rate": 1.978907006391223e-06,
+      "loss": 0.5243,
+      "step": 10343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.222376002164524,
+      "learning_rate": 1.977406749598382e-06,
+      "loss": 0.5209,
+      "step": 10344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2206317361937802,
+      "learning_rate": 1.9759069993132405e-06,
+      "loss": 0.4998,
+      "step": 10345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1608321664498062,
+      "learning_rate": 1.9744077556304885e-06,
+      "loss": 0.4799,
+      "step": 10346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1636473152101807,
+      "learning_rate": 1.9729090186447853e-06,
+      "loss": 0.4657,
+      "step": 10347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2029562992869192,
+      "learning_rate": 1.9714107884507474e-06,
+      "loss": 0.5271,
+      "step": 10348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2204066534511777,
+      "learning_rate": 1.9699130651429676e-06,
+      "loss": 0.4997,
+      "step": 10349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1497398828640724,
+      "learning_rate": 1.9684158488160065e-06,
+      "loss": 0.5064,
+      "step": 10350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1631250382904588,
+      "learning_rate": 1.9669191395643906e-06,
+      "loss": 0.4446,
+      "step": 10351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1349997615393839,
+      "learning_rate": 1.965422937482616e-06,
+      "loss": 0.4859,
+      "step": 10352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0774769701344173,
+      "learning_rate": 1.9639272426651435e-06,
+      "loss": 0.4684,
+      "step": 10353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1835434402499982,
+      "learning_rate": 1.962432055206406e-06,
+      "loss": 0.4571,
+      "step": 10354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1445839514983547,
+      "learning_rate": 1.9609373752008e-06,
+      "loss": 0.5004,
+      "step": 10355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2275289142101145,
+      "learning_rate": 1.9594432027426925e-06,
+      "loss": 0.5227,
+      "step": 10356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2593028556662307,
+      "learning_rate": 1.9579495379264223e-06,
+      "loss": 0.4979,
+      "step": 10357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2159801699784512,
+      "learning_rate": 1.9564563808462867e-06,
+      "loss": 0.5156,
+      "step": 10358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2114540229036244,
+      "learning_rate": 1.9549637315965587e-06,
+      "loss": 0.5387,
+      "step": 10359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2060288399463288,
+      "learning_rate": 1.9534715902714775e-06,
+      "loss": 0.5223,
+      "step": 10360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2462118923899952,
+      "learning_rate": 1.9519799569652498e-06,
+      "loss": 0.5481,
+      "step": 10361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1920891977481711,
+      "learning_rate": 1.9504888317720515e-06,
+      "loss": 0.4916,
+      "step": 10362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.337029930649945,
+      "learning_rate": 1.948998214786022e-06,
+      "loss": 0.5658,
+      "step": 10363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1651741198959462,
+      "learning_rate": 1.9475081061012746e-06,
+      "loss": 0.4734,
+      "step": 10364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0222954821134806,
+      "learning_rate": 1.946018505811883e-06,
+      "loss": 0.4284,
+      "step": 10365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1425155231247648,
+      "learning_rate": 1.9445294140118965e-06,
+      "loss": 0.4887,
+      "step": 10366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1424013180341783,
+      "learning_rate": 1.9430408307953317e-06,
+      "loss": 0.5349,
+      "step": 10367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3395171365157932,
+      "learning_rate": 1.9415527562561655e-06,
+      "loss": 0.5672,
+      "step": 10368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1631960620520938,
+      "learning_rate": 1.9400651904883492e-06,
+      "loss": 0.5056,
+      "step": 10369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1770906476268992,
+      "learning_rate": 1.9385781335858014e-06,
+      "loss": 0.4884,
+      "step": 10370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2783768235575272,
+      "learning_rate": 1.937091585642408e-06,
+      "loss": 0.5395,
+      "step": 10371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1769348264652402,
+      "learning_rate": 1.935605546752023e-06,
+      "loss": 0.5312,
+      "step": 10372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2309911194333034,
+      "learning_rate": 1.934120017008465e-06,
+      "loss": 0.5031,
+      "step": 10373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2335654392581403,
+      "learning_rate": 1.932634996505528e-06,
+      "loss": 0.4825,
+      "step": 10374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1826541305807252,
+      "learning_rate": 1.931150485336962e-06,
+      "loss": 0.5027,
+      "step": 10375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1497857619876597,
+      "learning_rate": 1.9296664835964975e-06,
+      "loss": 0.4593,
+      "step": 10376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2023675689420044,
+      "learning_rate": 1.928182991377826e-06,
+      "loss": 0.5154,
+      "step": 10377
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1732740570307119,
+      "learning_rate": 1.926700008774606e-06,
+      "loss": 0.467,
+      "step": 10378
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2388419918389142,
+      "learning_rate": 1.9252175358804657e-06,
+      "loss": 0.4995,
+      "step": 10379
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2624740465725812,
+      "learning_rate": 1.9237355727890037e-06,
+      "loss": 0.525,
+      "step": 10380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2348135639890256,
+      "learning_rate": 1.922254119593784e-06,
+      "loss": 0.4428,
+      "step": 10381
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1873731545412778,
+      "learning_rate": 1.9207731763883388e-06,
+      "loss": 0.515,
+      "step": 10382
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.393612855741153,
+      "learning_rate": 1.9192927432661645e-06,
+      "loss": 0.5578,
+      "step": 10383
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1886440588497955,
+      "learning_rate": 1.9178128203207324e-06,
+      "loss": 0.5003,
+      "step": 10384
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2326494541819835,
+      "learning_rate": 1.916333407645472e-06,
+      "loss": 0.5637,
+      "step": 10385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.233964441325004,
+      "learning_rate": 1.914854505333791e-06,
+      "loss": 0.4889,
+      "step": 10386
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2268538402378648,
+      "learning_rate": 1.9133761134790618e-06,
+      "loss": 0.5438,
+      "step": 10387
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0855906090260632,
+      "learning_rate": 1.9118982321746173e-06,
+      "loss": 0.4048,
+      "step": 10388
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2265365014692144,
+      "learning_rate": 1.9104208615137654e-06,
+      "loss": 0.4727,
+      "step": 10389
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0774274034228972,
+      "learning_rate": 1.908944001589782e-06,
+      "loss": 0.469,
+      "step": 10390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.158155572182002,
+      "learning_rate": 1.907467652495909e-06,
+      "loss": 0.4746,
+      "step": 10391
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.191380384821055,
+      "learning_rate": 1.9059918143253564e-06,
+      "loss": 0.4869,
+      "step": 10392
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2269686372754622,
+      "learning_rate": 1.9045164871713007e-06,
+      "loss": 0.4924,
+      "step": 10393
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1213116583568241,
+      "learning_rate": 1.903041671126884e-06,
+      "loss": 0.4862,
+      "step": 10394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2229470099216209,
+      "learning_rate": 1.9015673662852207e-06,
+      "loss": 0.5175,
+      "step": 10395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3604078205905663,
+      "learning_rate": 1.900093572739392e-06,
+      "loss": 0.5074,
+      "step": 10396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1943786064504713,
+      "learning_rate": 1.898620290582447e-06,
+      "loss": 0.474,
+      "step": 10397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2919564332289357,
+      "learning_rate": 1.897147519907403e-06,
+      "loss": 0.5327,
+      "step": 10398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.187736387315714,
+      "learning_rate": 1.8956752608072382e-06,
+      "loss": 0.5016,
+      "step": 10399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1687796339062262,
+      "learning_rate": 1.894203513374907e-06,
+      "loss": 0.5086,
+      "step": 10400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3468612316162474,
+      "learning_rate": 1.8927322777033285e-06,
+      "loss": 0.5067,
+      "step": 10401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2436114612654876,
+      "learning_rate": 1.8912615538853919e-06,
+      "loss": 0.5292,
+      "step": 10402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2027217139543736,
+      "learning_rate": 1.8897913420139492e-06,
+      "loss": 0.4489,
+      "step": 10403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1082299999357392,
+      "learning_rate": 1.8883216421818196e-06,
+      "loss": 0.4796,
+      "step": 10404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1507077257710712,
+      "learning_rate": 1.8868524544817957e-06,
+      "loss": 0.5064,
+      "step": 10405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1087962507989588,
+      "learning_rate": 1.8853837790066343e-06,
+      "loss": 0.511,
+      "step": 10406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2132909240140473,
+      "learning_rate": 1.8839156158490612e-06,
+      "loss": 0.5222,
+      "step": 10407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2021377277904344,
+      "learning_rate": 1.8824479651017712e-06,
+      "loss": 0.4608,
+      "step": 10408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.183529137588327,
+      "learning_rate": 1.8809808268574192e-06,
+      "loss": 0.5101,
+      "step": 10409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1541069894412248,
+      "learning_rate": 1.8795142012086364e-06,
+      "loss": 0.4803,
+      "step": 10410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2045538208410849,
+      "learning_rate": 1.8780480882480189e-06,
+      "loss": 0.5278,
+      "step": 10411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3228246952236407,
+      "learning_rate": 1.8765824880681317e-06,
+      "loss": 0.5211,
+      "step": 10412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2195136783172147,
+      "learning_rate": 1.8751174007615026e-06,
+      "loss": 0.4967,
+      "step": 10413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2479047858554162,
+      "learning_rate": 1.8736528264206289e-06,
+      "loss": 0.5097,
+      "step": 10414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.158638832002312,
+      "learning_rate": 1.872188765137979e-06,
+      "loss": 0.5247,
+      "step": 10415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.10435893825904,
+      "learning_rate": 1.8707252170059864e-06,
+      "loss": 0.4589,
+      "step": 10416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.214204184415717,
+      "learning_rate": 1.869262182117052e-06,
+      "loss": 0.4974,
+      "step": 10417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.146734571044995,
+      "learning_rate": 1.8677996605635473e-06,
+      "loss": 0.4902,
+      "step": 10418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1523582845514786,
+      "learning_rate": 1.866337652437805e-06,
+      "loss": 0.5315,
+      "step": 10419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1235626893729556,
+      "learning_rate": 1.8648761578321296e-06,
+      "loss": 0.4829,
+      "step": 10420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1653543766647423,
+      "learning_rate": 1.8634151768387954e-06,
+      "loss": 0.4622,
+      "step": 10421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0817316515014181,
+      "learning_rate": 1.8619547095500423e-06,
+      "loss": 0.4459,
+      "step": 10422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.241798531914272,
+      "learning_rate": 1.8604947560580756e-06,
+      "loss": 0.4962,
+      "step": 10423
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1386425609970265,
+      "learning_rate": 1.8590353164550656e-06,
+      "loss": 0.5111,
+      "step": 10424
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.244623207387527,
+      "learning_rate": 1.857576390833159e-06,
+      "loss": 0.5167,
+      "step": 10425
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.172029001925191,
+      "learning_rate": 1.8561179792844642e-06,
+      "loss": 0.5243,
+      "step": 10426
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2720301053185297,
+      "learning_rate": 1.8546600819010575e-06,
+      "loss": 0.5693,
+      "step": 10427
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.156249742250156,
+      "learning_rate": 1.8532026987749874e-06,
+      "loss": 0.4858,
+      "step": 10428
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2187584607735416,
+      "learning_rate": 1.8517458299982604e-06,
+      "loss": 0.519,
+      "step": 10429
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2533085429414896,
+      "learning_rate": 1.8502894756628587e-06,
+      "loss": 0.5415,
+      "step": 10430
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1305972307716194,
+      "learning_rate": 1.8488336358607296e-06,
+      "loss": 0.453,
+      "step": 10431
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2000558522260178,
+      "learning_rate": 1.8473783106837896e-06,
+      "loss": 0.4781,
+      "step": 10432
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1443368271399246,
+      "learning_rate": 1.8459235002239183e-06,
+      "loss": 0.4921,
+      "step": 10433
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2608846263364868,
+      "learning_rate": 1.844469204572964e-06,
+      "loss": 0.557,
+      "step": 10434
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1178104824424413,
+      "learning_rate": 1.843015423822746e-06,
+      "loss": 0.4885,
+      "step": 10435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1748830209493364,
+      "learning_rate": 1.841562158065049e-06,
+      "loss": 0.5095,
+      "step": 10436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.165367367987973,
+      "learning_rate": 1.8401094073916237e-06,
+      "loss": 0.4664,
+      "step": 10437
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.295025483326214,
+      "learning_rate": 1.8386571718941947e-06,
+      "loss": 0.5833,
+      "step": 10438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1498459464009567,
+      "learning_rate": 1.8372054516644422e-06,
+      "loss": 0.4455,
+      "step": 10439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1600520539936359,
+      "learning_rate": 1.8357542467940249e-06,
+      "loss": 0.4456,
+      "step": 10440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1858258241284776,
+      "learning_rate": 1.8343035573745637e-06,
+      "loss": 0.4609,
+      "step": 10441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2969841968570677,
+      "learning_rate": 1.8328533834976503e-06,
+      "loss": 0.538,
+      "step": 10442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.247649653446576,
+      "learning_rate": 1.83140372525484e-06,
+      "loss": 0.5145,
+      "step": 10443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1970741903024715,
+      "learning_rate": 1.8299545827376552e-06,
+      "loss": 0.5031,
+      "step": 10444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1042433778038725,
+      "learning_rate": 1.8285059560375883e-06,
+      "loss": 0.5108,
+      "step": 10445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.154651619607953,
+      "learning_rate": 1.8270578452461007e-06,
+      "loss": 0.4967,
+      "step": 10446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1493458691172969,
+      "learning_rate": 1.825610250454618e-06,
+      "loss": 0.5224,
+      "step": 10447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2318701620021442,
+      "learning_rate": 1.824163171754536e-06,
+      "loss": 0.4985,
+      "step": 10448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0756030232180962,
+      "learning_rate": 1.8227166092372138e-06,
+      "loss": 0.4431,
+      "step": 10449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1653508986481198,
+      "learning_rate": 1.8212705629939798e-06,
+      "loss": 0.5148,
+      "step": 10450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4137273992052395,
+      "learning_rate": 1.8198250331161327e-06,
+      "loss": 0.5678,
+      "step": 10451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2140237180978994,
+      "learning_rate": 1.8183800196949375e-06,
+      "loss": 0.5141,
+      "step": 10452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1900106792611889,
+      "learning_rate": 1.8169355228216211e-06,
+      "loss": 0.524,
+      "step": 10453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2498308544158938,
+      "learning_rate": 1.8154915425873865e-06,
+      "loss": 0.4813,
+      "step": 10454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2530844303708175,
+      "learning_rate": 1.8140480790833958e-06,
+      "loss": 0.5208,
+      "step": 10455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1372842856635434,
+      "learning_rate": 1.8126051324007821e-06,
+      "loss": 0.4889,
+      "step": 10456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1247584825455763,
+      "learning_rate": 1.8111627026306488e-06,
+      "loss": 0.4374,
+      "step": 10457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2168712807526494,
+      "learning_rate": 1.8097207898640633e-06,
+      "loss": 0.4463,
+      "step": 10458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.215735693748371,
+      "learning_rate": 1.808279394192063e-06,
+      "loss": 0.5298,
+      "step": 10459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3411277646325948,
+      "learning_rate": 1.8068385157056446e-06,
+      "loss": 0.5594,
+      "step": 10460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.203137880727011,
+      "learning_rate": 1.8053981544957832e-06,
+      "loss": 0.5018,
+      "step": 10461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3381786175693573,
+      "learning_rate": 1.8039583106534164e-06,
+      "loss": 0.5225,
+      "step": 10462
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1489099062930568,
+      "learning_rate": 1.8025189842694458e-06,
+      "loss": 0.5054,
+      "step": 10463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.280028575116321,
+      "learning_rate": 1.8010801754347473e-06,
+      "loss": 0.5046,
+      "step": 10464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1998320760058416,
+      "learning_rate": 1.7996418842401552e-06,
+      "loss": 0.5023,
+      "step": 10465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.277502440881357,
+      "learning_rate": 1.7982041107764803e-06,
+      "loss": 0.5216,
+      "step": 10466
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2627140052633197,
+      "learning_rate": 1.796766855134494e-06,
+      "loss": 0.5333,
+      "step": 10467
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3363706990641024,
+      "learning_rate": 1.795330117404941e-06,
+      "loss": 0.4937,
+      "step": 10468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1386407811939736,
+      "learning_rate": 1.7938938976785302e-06,
+      "loss": 0.5339,
+      "step": 10469
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1885433131258651,
+      "learning_rate": 1.792458196045932e-06,
+      "loss": 0.4912,
+      "step": 10470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1723792453554376,
+      "learning_rate": 1.7910230125977945e-06,
+      "loss": 0.483,
+      "step": 10471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3337188948031438,
+      "learning_rate": 1.789588347424729e-06,
+      "loss": 0.5345,
+      "step": 10472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1924109553234687,
+      "learning_rate": 1.7881542006173091e-06,
+      "loss": 0.5484,
+      "step": 10473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2019992862896063,
+      "learning_rate": 1.786720572266084e-06,
+      "loss": 0.4812,
+      "step": 10474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1503896923741443,
+      "learning_rate": 1.7852874624615624e-06,
+      "loss": 0.5159,
+      "step": 10475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1572777072855462,
+      "learning_rate": 1.783854871294225e-06,
+      "loss": 0.4836,
+      "step": 10476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1532732302287767,
+      "learning_rate": 1.7824227988545194e-06,
+      "loss": 0.509,
+      "step": 10477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3525361980311499,
+      "learning_rate": 1.7809912452328592e-06,
+      "loss": 0.5405,
+      "step": 10478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.25077342901221,
+      "learning_rate": 1.7795602105196297e-06,
+      "loss": 0.543,
+      "step": 10479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1658554094872522,
+      "learning_rate": 1.778129694805173e-06,
+      "loss": 0.4991,
+      "step": 10480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.216405128865592,
+      "learning_rate": 1.776699698179808e-06,
+      "loss": 0.4853,
+      "step": 10481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.162459735788342,
+      "learning_rate": 1.7752702207338202e-06,
+      "loss": 0.5001,
+      "step": 10482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1763861231425468,
+      "learning_rate": 1.7738412625574542e-06,
+      "loss": 0.4947,
+      "step": 10483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3161269437790237,
+      "learning_rate": 1.7724128237409344e-06,
+      "loss": 0.5357,
+      "step": 10484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.297538541614623,
+      "learning_rate": 1.7709849043744387e-06,
+      "loss": 0.5203,
+      "step": 10485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.199755321830302,
+      "learning_rate": 1.7695575045481218e-06,
+      "loss": 0.4795,
+      "step": 10486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1301971763142185,
+      "learning_rate": 1.7681306243521035e-06,
+      "loss": 0.4911,
+      "step": 10487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0538742320723389,
+      "learning_rate": 1.7667042638764697e-06,
+      "loss": 0.4589,
+      "step": 10488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0407850858339627,
+      "learning_rate": 1.7652784232112763e-06,
+      "loss": 0.4434,
+      "step": 10489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1943022005021457,
+      "learning_rate": 1.76385310244654e-06,
+      "loss": 0.5133,
+      "step": 10490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3291034908687103,
+      "learning_rate": 1.762428301672251e-06,
+      "loss": 0.5734,
+      "step": 10491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3266406122761054,
+      "learning_rate": 1.761004020978363e-06,
+      "loss": 0.6024,
+      "step": 10492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1645543640209564,
+      "learning_rate": 1.7595802604547974e-06,
+      "loss": 0.496,
+      "step": 10493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2184970299735831,
+      "learning_rate": 1.7581570201914478e-06,
+      "loss": 0.5359,
+      "step": 10494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1513206983762794,
+      "learning_rate": 1.7567343002781656e-06,
+      "loss": 0.5129,
+      "step": 10495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.330582264097016,
+      "learning_rate": 1.7553121008047768e-06,
+      "loss": 0.5321,
+      "step": 10496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.209381827007413,
+      "learning_rate": 1.753890421861072e-06,
+      "loss": 0.5503,
+      "step": 10497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2934417781312084,
+      "learning_rate": 1.752469263536809e-06,
+      "loss": 0.5061,
+      "step": 10498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2345777900323305,
+      "learning_rate": 1.7510486259217151e-06,
+      "loss": 0.5093,
+      "step": 10499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2739734574941015,
+      "learning_rate": 1.7496285091054788e-06,
+      "loss": 0.5533,
+      "step": 10500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1736133144039462,
+      "learning_rate": 1.7482089131777635e-06,
+      "loss": 0.4895,
+      "step": 10501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.244689484981838,
+      "learning_rate": 1.74678983822819e-06,
+      "loss": 0.4912,
+      "step": 10502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3912718479329513,
+      "learning_rate": 1.7453712843463554e-06,
+      "loss": 0.5747,
+      "step": 10503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2169156085479538,
+      "learning_rate": 1.7439532516218226e-06,
+      "loss": 0.5029,
+      "step": 10504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1270013068359976,
+      "learning_rate": 1.7425357401441134e-06,
+      "loss": 0.4715,
+      "step": 10505
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.253966570677824,
+      "learning_rate": 1.7411187500027272e-06,
+      "loss": 0.5125,
+      "step": 10506
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2157143175119676,
+      "learning_rate": 1.7397022812871234e-06,
+      "loss": 0.5049,
+      "step": 10507
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2118043811480281,
+      "learning_rate": 1.7382863340867316e-06,
+      "loss": 0.5172,
+      "step": 10508
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3806981878883626,
+      "learning_rate": 1.7368709084909496e-06,
+      "loss": 0.5133,
+      "step": 10509
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2219046333017223,
+      "learning_rate": 1.735456004589141e-06,
+      "loss": 0.4581,
+      "step": 10510
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.217999594091714,
+      "learning_rate": 1.7340416224706346e-06,
+      "loss": 0.4843,
+      "step": 10511
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.162585044129144,
+      "learning_rate": 1.7326277622247245e-06,
+      "loss": 0.5077,
+      "step": 10512
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.092069042587265,
+      "learning_rate": 1.7312144239406781e-06,
+      "loss": 0.4631,
+      "step": 10513
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3889924228013881,
+      "learning_rate": 1.7298016077077273e-06,
+      "loss": 0.5579,
+      "step": 10514
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1529110821220014,
+      "learning_rate": 1.728389313615071e-06,
+      "loss": 0.4757,
+      "step": 10515
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.31379771791104,
+      "learning_rate": 1.7269775417518708e-06,
+      "loss": 0.5394,
+      "step": 10516
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2688102670270993,
+      "learning_rate": 1.7255662922072613e-06,
+      "loss": 0.505,
+      "step": 10517
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2353242531875206,
+      "learning_rate": 1.7241555650703433e-06,
+      "loss": 0.5218,
+      "step": 10518
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.267278321347615,
+      "learning_rate": 1.7227453604301814e-06,
+      "loss": 0.4969,
+      "step": 10519
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1978121781043962,
+      "learning_rate": 1.7213356783758128e-06,
+      "loss": 0.4724,
+      "step": 10520
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.210791006763986,
+      "learning_rate": 1.7199265189962345e-06,
+      "loss": 0.5214,
+      "step": 10521
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.239744314858782,
+      "learning_rate": 1.7185178823804127e-06,
+      "loss": 0.553,
+      "step": 10522
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2265063716413458,
+      "learning_rate": 1.7171097686172832e-06,
+      "loss": 0.5414,
+      "step": 10523
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1923382225103458,
+      "learning_rate": 1.7157021777957494e-06,
+      "loss": 0.4642,
+      "step": 10524
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.131043308828589,
+      "learning_rate": 1.7142951100046802e-06,
+      "loss": 0.4543,
+      "step": 10525
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1441874329436754,
+      "learning_rate": 1.7128885653329065e-06,
+      "loss": 0.4569,
+      "step": 10526
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1432271043531488,
+      "learning_rate": 1.711482543869234e-06,
+      "loss": 0.5425,
+      "step": 10527
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0816328506493416,
+      "learning_rate": 1.7100770457024307e-06,
+      "loss": 0.5059,
+      "step": 10528
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4127508582100923,
+      "learning_rate": 1.7086720709212357e-06,
+      "loss": 0.5717,
+      "step": 10529
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.156142100892866,
+      "learning_rate": 1.7072676196143512e-06,
+      "loss": 0.5183,
+      "step": 10530
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1123369997604466,
+      "learning_rate": 1.7058636918704474e-06,
+      "loss": 0.4447,
+      "step": 10531
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.188770768699168,
+      "learning_rate": 1.704460287778159e-06,
+      "loss": 0.4777,
+      "step": 10532
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3072897263876002,
+      "learning_rate": 1.7030574074260908e-06,
+      "loss": 0.5636,
+      "step": 10533
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.143892650359272,
+      "learning_rate": 1.7016550509028162e-06,
+      "loss": 0.4896,
+      "step": 10534
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.130228977044135,
+      "learning_rate": 1.7002532182968734e-06,
+      "loss": 0.4782,
+      "step": 10535
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1603531590084593,
+      "learning_rate": 1.6988519096967647e-06,
+      "loss": 0.5064,
+      "step": 10536
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1335112948567967,
+      "learning_rate": 1.697451125190962e-06,
+      "loss": 0.508,
+      "step": 10537
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2376035646971673,
+      "learning_rate": 1.6960508648679064e-06,
+      "loss": 0.5099,
+      "step": 10538
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2352067101031956,
+      "learning_rate": 1.6946511288160017e-06,
+      "loss": 0.5089,
+      "step": 10539
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1461286626593559,
+      "learning_rate": 1.693251917123624e-06,
+      "loss": 0.4921,
+      "step": 10540
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3278851460806542,
+      "learning_rate": 1.6918532298791102e-06,
+      "loss": 0.5413,
+      "step": 10541
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2688491162450473,
+      "learning_rate": 1.6904550671707632e-06,
+      "loss": 0.4886,
+      "step": 10542
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3579935969312724,
+      "learning_rate": 1.68905742908686e-06,
+      "loss": 0.5718,
+      "step": 10543
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2583445495880312,
+      "learning_rate": 1.6876603157156402e-06,
+      "loss": 0.5128,
+      "step": 10544
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1596382603102533,
+      "learning_rate": 1.686263727145313e-06,
+      "loss": 0.4761,
+      "step": 10545
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4515619382662144,
+      "learning_rate": 1.684867663464047e-06,
+      "loss": 0.4988,
+      "step": 10546
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1546457863836639,
+      "learning_rate": 1.6834721247599871e-06,
+      "loss": 0.5124,
+      "step": 10547
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.371756542887517,
+      "learning_rate": 1.6820771111212386e-06,
+      "loss": 0.5701,
+      "step": 10548
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2650184590798446,
+      "learning_rate": 1.6806826226358775e-06,
+      "loss": 0.4859,
+      "step": 10549
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.297595134400607,
+      "learning_rate": 1.6792886593919454e-06,
+      "loss": 0.5629,
+      "step": 10550
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1903674490401843,
+      "learning_rate": 1.6778952214774513e-06,
+      "loss": 0.4868,
+      "step": 10551
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2320864261712654,
+      "learning_rate": 1.6765023089803645e-06,
+      "loss": 0.4971,
+      "step": 10552
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2018654410408118,
+      "learning_rate": 1.6751099219886314e-06,
+      "loss": 0.4749,
+      "step": 10553
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1420793536920715,
+      "learning_rate": 1.6737180605901592e-06,
+      "loss": 0.4877,
+      "step": 10554
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2102358354142642,
+      "learning_rate": 1.672326724872827e-06,
+      "loss": 0.4966,
+      "step": 10555
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1361701401783366,
+      "learning_rate": 1.6709359149244708e-06,
+      "loss": 0.4591,
+      "step": 10556
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.195401406565811,
+      "learning_rate": 1.6695456308329027e-06,
+      "loss": 0.4591,
+      "step": 10557
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2573925288634176,
+      "learning_rate": 1.6681558726858983e-06,
+      "loss": 0.5205,
+      "step": 10558
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2768040721461484,
+      "learning_rate": 1.6667666405712002e-06,
+      "loss": 0.4884,
+      "step": 10559
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.193325860135069,
+      "learning_rate": 1.6653779345765209e-06,
+      "loss": 0.5381,
+      "step": 10560
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0951323356961833,
+      "learning_rate": 1.6639897547895334e-06,
+      "loss": 0.4846,
+      "step": 10561
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2313968125547885,
+      "learning_rate": 1.662602101297879e-06,
+      "loss": 0.4823,
+      "step": 10562
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1681266856500292,
+      "learning_rate": 1.661214974189168e-06,
+      "loss": 0.4634,
+      "step": 10563
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2898688366470343,
+      "learning_rate": 1.6598283735509791e-06,
+      "loss": 0.5213,
+      "step": 10564
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2660283988159908,
+      "learning_rate": 1.6584422994708539e-06,
+      "loss": 0.5922,
+      "step": 10565
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2480377053653755,
+      "learning_rate": 1.6570567520363058e-06,
+      "loss": 0.5292,
+      "step": 10566
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1522369335258873,
+      "learning_rate": 1.6556717313348058e-06,
+      "loss": 0.4968,
+      "step": 10567
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2200183626262613,
+      "learning_rate": 1.654287237453801e-06,
+      "loss": 0.5661,
+      "step": 10568
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1331459343938992,
+      "learning_rate": 1.6529032704807012e-06,
+      "loss": 0.4938,
+      "step": 10569
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2208820669626403,
+      "learning_rate": 1.6515198305028868e-06,
+      "loss": 0.5376,
+      "step": 10570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.313192139642635,
+      "learning_rate": 1.6501369176076964e-06,
+      "loss": 0.5588,
+      "step": 10571
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.126480452979602,
+      "learning_rate": 1.6487545318824405e-06,
+      "loss": 0.4844,
+      "step": 10572
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0932797102067502,
+      "learning_rate": 1.6473726734143969e-06,
+      "loss": 0.4108,
+      "step": 10573
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1992780937894623,
+      "learning_rate": 1.6459913422908113e-06,
+      "loss": 0.5026,
+      "step": 10574
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1143017826214296,
+      "learning_rate": 1.6446105385988932e-06,
+      "loss": 0.479,
+      "step": 10575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0995762550691859,
+      "learning_rate": 1.6432302624258211e-06,
+      "loss": 0.4681,
+      "step": 10576
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1709815115963567,
+      "learning_rate": 1.641850513858737e-06,
+      "loss": 0.4255,
+      "step": 10577
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1894593387435193,
+      "learning_rate": 1.640471292984751e-06,
+      "loss": 0.5529,
+      "step": 10578
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1155630185302747,
+      "learning_rate": 1.6390925998909423e-06,
+      "loss": 0.4771,
+      "step": 10579
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2387560586944755,
+      "learning_rate": 1.637714434664357e-06,
+      "loss": 0.5498,
+      "step": 10580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.172648314589578,
+      "learning_rate": 1.6363367973920031e-06,
+      "loss": 0.4916,
+      "step": 10581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1733047918258428,
+      "learning_rate": 1.6349596881608555e-06,
+      "loss": 0.5129,
+      "step": 10582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0808998555280984,
+      "learning_rate": 1.6335831070578612e-06,
+      "loss": 0.4442,
+      "step": 10583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.420055189000459,
+      "learning_rate": 1.6322070541699298e-06,
+      "loss": 0.5632,
+      "step": 10584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4057960625309385,
+      "learning_rate": 1.6308315295839395e-06,
+      "loss": 0.5387,
+      "step": 10585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0657794493674142,
+      "learning_rate": 1.6294565333867362e-06,
+      "loss": 0.4544,
+      "step": 10586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0917078299199416,
+      "learning_rate": 1.6280820656651252e-06,
+      "loss": 0.4467,
+      "step": 10587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1970082639601753,
+      "learning_rate": 1.6267081265058882e-06,
+      "loss": 0.5093,
+      "step": 10588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0844817249205754,
+      "learning_rate": 1.6253347159957666e-06,
+      "loss": 0.4324,
+      "step": 10589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3469163273181914,
+      "learning_rate": 1.6239618342214746e-06,
+      "loss": 0.4984,
+      "step": 10590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1597337563576289,
+      "learning_rate": 1.6225894812696875e-06,
+      "loss": 0.5273,
+      "step": 10591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1174359345473404,
+      "learning_rate": 1.6212176572270445e-06,
+      "loss": 0.5066,
+      "step": 10592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2679483728187548,
+      "learning_rate": 1.6198463621801607e-06,
+      "loss": 0.5363,
+      "step": 10593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1787600376676683,
+      "learning_rate": 1.6184755962156118e-06,
+      "loss": 0.4586,
+      "step": 10594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1610539107147957,
+      "learning_rate": 1.6171053594199403e-06,
+      "loss": 0.4825,
+      "step": 10595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1591404536227137,
+      "learning_rate": 1.6157356518796608e-06,
+      "loss": 0.4964,
+      "step": 10596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2739170786222964,
+      "learning_rate": 1.6143664736812449e-06,
+      "loss": 0.5388,
+      "step": 10597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1487295596822786,
+      "learning_rate": 1.6129978249111366e-06,
+      "loss": 0.5308,
+      "step": 10598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.053988585426585,
+      "learning_rate": 1.6116297056557507e-06,
+      "loss": 0.4454,
+      "step": 10599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.140154493303633,
+      "learning_rate": 1.6102621160014563e-06,
+      "loss": 0.4721,
+      "step": 10600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0979149130144294,
+      "learning_rate": 1.6088950560346017e-06,
+      "loss": 0.5351,
+      "step": 10601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3602556029114805,
+      "learning_rate": 1.607528525841493e-06,
+      "loss": 0.5462,
+      "step": 10602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2487420905304725,
+      "learning_rate": 1.6061625255084079e-06,
+      "loss": 0.4939,
+      "step": 10603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.264048925451849,
+      "learning_rate": 1.6047970551215898e-06,
+      "loss": 0.5182,
+      "step": 10604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3149038735128054,
+      "learning_rate": 1.603432114767246e-06,
+      "loss": 0.5147,
+      "step": 10605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1976032123405373,
+      "learning_rate": 1.6020677045315558e-06,
+      "loss": 0.4532,
+      "step": 10606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2281628998399992,
+      "learning_rate": 1.600703824500658e-06,
+      "loss": 0.5406,
+      "step": 10607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2557731824280873,
+      "learning_rate": 1.5993404747606612e-06,
+      "loss": 0.5092,
+      "step": 10608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2202247597855143,
+      "learning_rate": 1.5979776553976444e-06,
+      "loss": 0.5009,
+      "step": 10609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.260890109886381,
+      "learning_rate": 1.5966153664976447e-06,
+      "loss": 0.5499,
+      "step": 10610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1915798373477975,
+      "learning_rate": 1.5952536081466752e-06,
+      "loss": 0.4717,
+      "step": 10611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2130895872219338,
+      "learning_rate": 1.593892380430706e-06,
+      "loss": 0.4822,
+      "step": 10612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1936408427396823,
+      "learning_rate": 1.5925316834356797e-06,
+      "loss": 0.4967,
+      "step": 10613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.113920918118061,
+      "learning_rate": 1.5911715172475062e-06,
+      "loss": 0.4447,
+      "step": 10614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.354397397069576,
+      "learning_rate": 1.589811881952058e-06,
+      "loss": 0.5498,
+      "step": 10615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1929171128205,
+      "learning_rate": 1.5884527776351765e-06,
+      "loss": 0.4722,
+      "step": 10616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1839754603297938,
+      "learning_rate": 1.587094204382672e-06,
+      "loss": 0.5038,
+      "step": 10617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1277828236536438,
+      "learning_rate": 1.5857361622803124e-06,
+      "loss": 0.4634,
+      "step": 10618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2086732430136258,
+      "learning_rate": 1.5843786514138438e-06,
+      "loss": 0.5452,
+      "step": 10619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.243875041024383,
+      "learning_rate": 1.5830216718689674e-06,
+      "loss": 0.4856,
+      "step": 10620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1107088586063996,
+      "learning_rate": 1.5816652237313579e-06,
+      "loss": 0.4601,
+      "step": 10621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.141746753241139,
+      "learning_rate": 1.5803093070866582e-06,
+      "loss": 0.4985,
+      "step": 10622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.138993494274794,
+      "learning_rate": 1.5789539220204698e-06,
+      "loss": 0.4817,
+      "step": 10623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2651400637830874,
+      "learning_rate": 1.5775990686183672e-06,
+      "loss": 0.5021,
+      "step": 10624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.303035626902938,
+      "learning_rate": 1.5762447469658892e-06,
+      "loss": 0.5787,
+      "step": 10625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.141180399666159,
+      "learning_rate": 1.5748909571485415e-06,
+      "loss": 0.4397,
+      "step": 10626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2237553637877017,
+      "learning_rate": 1.573537699251797e-06,
+      "loss": 0.4826,
+      "step": 10627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1775421947362295,
+      "learning_rate": 1.5721849733610905e-06,
+      "loss": 0.5236,
+      "step": 10628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0834415210597546,
+      "learning_rate": 1.5708327795618317e-06,
+      "loss": 0.4447,
+      "step": 10629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.165043001895713,
+      "learning_rate": 1.5694811179393853e-06,
+      "loss": 0.5152,
+      "step": 10630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2859387551877466,
+      "learning_rate": 1.5681299885790912e-06,
+      "loss": 0.4722,
+      "step": 10631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2709186191811892,
+      "learning_rate": 1.5667793915662566e-06,
+      "loss": 0.5463,
+      "step": 10632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.161564803553512,
+      "learning_rate": 1.5654293269861464e-06,
+      "loss": 0.4952,
+      "step": 10633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.295246480586887,
+      "learning_rate": 1.5640797949239983e-06,
+      "loss": 0.5319,
+      "step": 10634
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2390392883358687,
+      "learning_rate": 1.5627307954650174e-06,
+      "loss": 0.4864,
+      "step": 10635
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1896794045976744,
+      "learning_rate": 1.5613823286943718e-06,
+      "loss": 0.5396,
+      "step": 10636
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2407145374187092,
+      "learning_rate": 1.5600343946971997e-06,
+      "loss": 0.542,
+      "step": 10637
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1586477831679753,
+      "learning_rate": 1.558686993558599e-06,
+      "loss": 0.493,
+      "step": 10638
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1317715828137198,
+      "learning_rate": 1.557340125363641e-06,
+      "loss": 0.4916,
+      "step": 10639
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.250607247672225,
+      "learning_rate": 1.5559937901973575e-06,
+      "loss": 0.5091,
+      "step": 10640
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2454636275081175,
+      "learning_rate": 1.554647988144752e-06,
+      "loss": 0.4796,
+      "step": 10641
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1786074715424588,
+      "learning_rate": 1.5533027192907924e-06,
+      "loss": 0.5,
+      "step": 10642
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3496931963670329,
+      "learning_rate": 1.5519579837204103e-06,
+      "loss": 0.515,
+      "step": 10643
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.155420314967855,
+      "learning_rate": 1.5506137815185063e-06,
+      "loss": 0.4891,
+      "step": 10644
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1867319434049721,
+      "learning_rate": 1.5492701127699472e-06,
+      "loss": 0.4803,
+      "step": 10645
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0859875564564818,
+      "learning_rate": 1.5479269775595652e-06,
+      "loss": 0.4627,
+      "step": 10646
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3417575169639346,
+      "learning_rate": 1.5465843759721633e-06,
+      "loss": 0.4812,
+      "step": 10647
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0905783307165362,
+      "learning_rate": 1.5452423080925017e-06,
+      "loss": 0.4835,
+      "step": 10648
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1912132732257752,
+      "learning_rate": 1.5439007740053158e-06,
+      "loss": 0.5001,
+      "step": 10649
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1660853481692872,
+      "learning_rate": 1.542559773795299e-06,
+      "loss": 0.5076,
+      "step": 10650
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2455843658633547,
+      "learning_rate": 1.5412193075471193e-06,
+      "loss": 0.5132,
+      "step": 10651
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2122231826191785,
+      "learning_rate": 1.5398793753454079e-06,
+      "loss": 0.5049,
+      "step": 10652
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2730598591847297,
+      "learning_rate": 1.5385399772747578e-06,
+      "loss": 0.5382,
+      "step": 10653
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2040212191394304,
+      "learning_rate": 1.537201113419735e-06,
+      "loss": 0.5187,
+      "step": 10654
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2002762595706624,
+      "learning_rate": 1.5358627838648676e-06,
+      "loss": 0.5191,
+      "step": 10655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1650171141943202,
+      "learning_rate": 1.5345249886946512e-06,
+      "loss": 0.5007,
+      "step": 10656
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1814721862289355,
+      "learning_rate": 1.5331877279935515e-06,
+      "loss": 0.5262,
+      "step": 10657
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.287374744063012,
+      "learning_rate": 1.5318510018459908e-06,
+      "loss": 0.5571,
+      "step": 10658
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0429881804867251,
+      "learning_rate": 1.5305148103363698e-06,
+      "loss": 0.4691,
+      "step": 10659
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2026368178732705,
+      "learning_rate": 1.5291791535490419e-06,
+      "loss": 0.5112,
+      "step": 10660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1618844463750302,
+      "learning_rate": 1.5278440315683385e-06,
+      "loss": 0.4974,
+      "step": 10661
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.279191247204793,
+      "learning_rate": 1.5265094444785544e-06,
+      "loss": 0.5216,
+      "step": 10662
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2477094167179268,
+      "learning_rate": 1.5251753923639435e-06,
+      "loss": 0.535,
+      "step": 10663
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1094570398217969,
+      "learning_rate": 1.5238418753087347e-06,
+      "loss": 0.4709,
+      "step": 10664
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.150298602272318,
+      "learning_rate": 1.52250889339712e-06,
+      "loss": 0.5148,
+      "step": 10665
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2391737841051351,
+      "learning_rate": 1.5211764467132562e-06,
+      "loss": 0.5111,
+      "step": 10666
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2618902698063414,
+      "learning_rate": 1.5198445353412705e-06,
+      "loss": 0.5339,
+      "step": 10667
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2754243668069645,
+      "learning_rate": 1.5185131593652492e-06,
+      "loss": 0.5474,
+      "step": 10668
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1565161733720979,
+      "learning_rate": 1.5171823188692537e-06,
+      "loss": 0.4839,
+      "step": 10669
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2418106755116487,
+      "learning_rate": 1.515852013937301e-06,
+      "loss": 0.51,
+      "step": 10670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1223652504326402,
+      "learning_rate": 1.5145222446533835e-06,
+      "loss": 0.5,
+      "step": 10671
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2221621163838106,
+      "learning_rate": 1.5131930111014558e-06,
+      "loss": 0.4724,
+      "step": 10672
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2141082107009922,
+      "learning_rate": 1.5118643133654421e-06,
+      "loss": 0.544,
+      "step": 10673
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2580655714118167,
+      "learning_rate": 1.5105361515292248e-06,
+      "loss": 0.4804,
+      "step": 10674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2292671270319728,
+      "learning_rate": 1.5092085256766597e-06,
+      "loss": 0.5168,
+      "step": 10675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0648835877924034,
+      "learning_rate": 1.5078814358915673e-06,
+      "loss": 0.4423,
+      "step": 10676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1668641343348982,
+      "learning_rate": 1.5065548822577336e-06,
+      "loss": 0.4548,
+      "step": 10677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1516953535320928,
+      "learning_rate": 1.5052288648589131e-06,
+      "loss": 0.4911,
+      "step": 10678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.219413625554132,
+      "learning_rate": 1.503903383778822e-06,
+      "loss": 0.5264,
+      "step": 10679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2380641903451444,
+      "learning_rate": 1.502578439101141e-06,
+      "loss": 0.5702,
+      "step": 10680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1643114271843449,
+      "learning_rate": 1.5012540309095247e-06,
+      "loss": 0.4866,
+      "step": 10681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.199663160653029,
+      "learning_rate": 1.4999301592875891e-06,
+      "loss": 0.529,
+      "step": 10682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1609746441861055,
+      "learning_rate": 1.4986068243189188e-06,
+      "loss": 0.4953,
+      "step": 10683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2624824031698814,
+      "learning_rate": 1.4972840260870603e-06,
+      "loss": 0.5146,
+      "step": 10684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2543424519040083,
+      "learning_rate": 1.4959617646755276e-06,
+      "loss": 0.5037,
+      "step": 10685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2434565940861557,
+      "learning_rate": 1.494640040167805e-06,
+      "loss": 0.5153,
+      "step": 10686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2467298647897496,
+      "learning_rate": 1.4933188526473385e-06,
+      "loss": 0.4857,
+      "step": 10687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.161222334122599,
+      "learning_rate": 1.4919982021975432e-06,
+      "loss": 0.4957,
+      "step": 10688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.101285588515358,
+      "learning_rate": 1.490678088901798e-06,
+      "loss": 0.467,
+      "step": 10689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1081466859437221,
+      "learning_rate": 1.489358512843444e-06,
+      "loss": 0.472,
+      "step": 10690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.142412900797478,
+      "learning_rate": 1.4880394741057968e-06,
+      "loss": 0.4944,
+      "step": 10691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.178900854301871,
+      "learning_rate": 1.4867209727721332e-06,
+      "loss": 0.4587,
+      "step": 10692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.22633194426983,
+      "learning_rate": 1.4854030089257e-06,
+      "loss": 0.5165,
+      "step": 10693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.154534123497548,
+      "learning_rate": 1.4840855826497013e-06,
+      "loss": 0.5071,
+      "step": 10694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2420594731218504,
+      "learning_rate": 1.4827686940273178e-06,
+      "loss": 0.4821,
+      "step": 10695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2937649583182145,
+      "learning_rate": 1.4814523431416882e-06,
+      "loss": 0.5263,
+      "step": 10696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2260895807480467,
+      "learning_rate": 1.4801365300759253e-06,
+      "loss": 0.4687,
+      "step": 10697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3067974914450806,
+      "learning_rate": 1.4788212549130964e-06,
+      "loss": 0.5591,
+      "step": 10698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3168350562730655,
+      "learning_rate": 1.4775065177362492e-06,
+      "loss": 0.5635,
+      "step": 10699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1561035888352202,
+      "learning_rate": 1.4761923186283822e-06,
+      "loss": 0.4237,
+      "step": 10700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1750204470052563,
+      "learning_rate": 1.4748786576724716e-06,
+      "loss": 0.4973,
+      "step": 10701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3664606832716801,
+      "learning_rate": 1.4735655349514555e-06,
+      "loss": 0.5479,
+      "step": 10702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.28754793837104,
+      "learning_rate": 1.4722529505482396e-06,
+      "loss": 0.5688,
+      "step": 10703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3797582312281547,
+      "learning_rate": 1.47094090454569e-06,
+      "loss": 0.5157,
+      "step": 10704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2582705119958655,
+      "learning_rate": 1.4696293970266463e-06,
+      "loss": 0.5462,
+      "step": 10705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1306954431388179,
+      "learning_rate": 1.4683184280739082e-06,
+      "loss": 0.4743,
+      "step": 10706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2099052705805882,
+      "learning_rate": 1.467007997770249e-06,
+      "loss": 0.493,
+      "step": 10707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1512392602309827,
+      "learning_rate": 1.4656981061983966e-06,
+      "loss": 0.5054,
+      "step": 10708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2088871979984055,
+      "learning_rate": 1.464388753441056e-06,
+      "loss": 0.5698,
+      "step": 10709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2783778959379315,
+      "learning_rate": 1.4630799395808893e-06,
+      "loss": 0.577,
+      "step": 10710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1232985239455786,
+      "learning_rate": 1.4617716647005298e-06,
+      "loss": 0.4855,
+      "step": 10711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2508608238169414,
+      "learning_rate": 1.4604639288825773e-06,
+      "loss": 0.4672,
+      "step": 10712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1214602200395811,
+      "learning_rate": 1.4591567322095978e-06,
+      "loss": 0.4392,
+      "step": 10713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1259987424467393,
+      "learning_rate": 1.4578500747641167e-06,
+      "loss": 0.5118,
+      "step": 10714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2225311020247238,
+      "learning_rate": 1.456543956628631e-06,
+      "loss": 0.4707,
+      "step": 10715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2100539398781158,
+      "learning_rate": 1.455238377885605e-06,
+      "loss": 0.4967,
+      "step": 10716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.137771224080233,
+      "learning_rate": 1.453933338617467e-06,
+      "loss": 0.5059,
+      "step": 10717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2057733493341938,
+      "learning_rate": 1.4526288389066068e-06,
+      "loss": 0.4884,
+      "step": 10718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2189737750292187,
+      "learning_rate": 1.4513248788353894e-06,
+      "loss": 0.4715,
+      "step": 10719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1831893985590594,
+      "learning_rate": 1.4500214584861349e-06,
+      "loss": 0.5097,
+      "step": 10720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.23139942637299,
+      "learning_rate": 1.4487185779411382e-06,
+      "loss": 0.4887,
+      "step": 10721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3156255563476547,
+      "learning_rate": 1.4474162372826562e-06,
+      "loss": 0.5304,
+      "step": 10722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2479036872880824,
+      "learning_rate": 1.446114436592916e-06,
+      "loss": 0.5598,
+      "step": 10723
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1843007811791368,
+      "learning_rate": 1.4448131759540996e-06,
+      "loss": 0.5057,
+      "step": 10724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2374767012281316,
+      "learning_rate": 1.4435124554483671e-06,
+      "loss": 0.519,
+      "step": 10725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.240530624451376,
+      "learning_rate": 1.4422122751578394e-06,
+      "loss": 0.5137,
+      "step": 10726
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2339787390406027,
+      "learning_rate": 1.4409126351646052e-06,
+      "loss": 0.5416,
+      "step": 10727
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2254550575192762,
+      "learning_rate": 1.4396135355507135e-06,
+      "loss": 0.5091,
+      "step": 10728
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1919957437951312,
+      "learning_rate": 1.4383149763981863e-06,
+      "loss": 0.5135,
+      "step": 10729
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1402930207704394,
+      "learning_rate": 1.4370169577890059e-06,
+      "loss": 0.5007,
+      "step": 10730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2649005653292626,
+      "learning_rate": 1.4357194798051244e-06,
+      "loss": 0.5077,
+      "step": 10731
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2354540876582532,
+      "learning_rate": 1.4344225425284565e-06,
+      "loss": 0.5043,
+      "step": 10732
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1919058330770496,
+      "learning_rate": 1.4331261460408874e-06,
+      "loss": 0.485,
+      "step": 10733
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2504249803994862,
+      "learning_rate": 1.4318302904242654e-06,
+      "loss": 0.4858,
+      "step": 10734
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1635475299702271,
+      "learning_rate": 1.4305349757604014e-06,
+      "loss": 0.5002,
+      "step": 10735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2957456336237656,
+      "learning_rate": 1.429240202131077e-06,
+      "loss": 0.4974,
+      "step": 10736
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.11815931827255,
+      "learning_rate": 1.4279459696180398e-06,
+      "loss": 0.4841,
+      "step": 10737
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3639096011608889,
+      "learning_rate": 1.4266522783029978e-06,
+      "loss": 0.5187,
+      "step": 10738
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2202920694631334,
+      "learning_rate": 1.4253591282676316e-06,
+      "loss": 0.5205,
+      "step": 10739
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1667114430872203,
+      "learning_rate": 1.424066519593581e-06,
+      "loss": 0.5079,
+      "step": 10740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1430497198830127,
+      "learning_rate": 1.4227744523624575e-06,
+      "loss": 0.4168,
+      "step": 10741
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3526177228508531,
+      "learning_rate": 1.421482926655835e-06,
+      "loss": 0.5473,
+      "step": 10742
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2059125932393542,
+      "learning_rate": 1.4201919425552557e-06,
+      "loss": 0.478,
+      "step": 10743
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1257928597502185,
+      "learning_rate": 1.4189015001422257e-06,
+      "loss": 0.4847,
+      "step": 10744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2357940725639691,
+      "learning_rate": 1.4176115994982153e-06,
+      "loss": 0.4943,
+      "step": 10745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.134541737306114,
+      "learning_rate": 1.416322240704664e-06,
+      "loss": 0.4747,
+      "step": 10746
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2373541601537699,
+      "learning_rate": 1.4150334238429776e-06,
+      "loss": 0.5243,
+      "step": 10747
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1749621608403984,
+      "learning_rate": 1.413745148994522e-06,
+      "loss": 0.4926,
+      "step": 10748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.211961030626305,
+      "learning_rate": 1.4124574162406356e-06,
+      "loss": 0.5169,
+      "step": 10749
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1651775984402475,
+      "learning_rate": 1.411170225662617e-06,
+      "loss": 0.4731,
+      "step": 10750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2407771327141737,
+      "learning_rate": 1.409883577341734e-06,
+      "loss": 0.5455,
+      "step": 10751
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.365969654997289,
+      "learning_rate": 1.4085974713592199e-06,
+      "loss": 0.5209,
+      "step": 10752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2404837770581318,
+      "learning_rate": 1.4073119077962738e-06,
+      "loss": 0.5274,
+      "step": 10753
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0377283350613595,
+      "learning_rate": 1.4060268867340621e-06,
+      "loss": 0.4271,
+      "step": 10754
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1906535788482806,
+      "learning_rate": 1.4047424082537086e-06,
+      "loss": 0.5017,
+      "step": 10755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2003521859407196,
+      "learning_rate": 1.4034584724363131e-06,
+      "loss": 0.5402,
+      "step": 10756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1219848387776727,
+      "learning_rate": 1.4021750793629397e-06,
+      "loss": 0.4741,
+      "step": 10757
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.196548222087757,
+      "learning_rate": 1.4008922291146087e-06,
+      "loss": 0.4337,
+      "step": 10758
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1567767721771123,
+      "learning_rate": 1.3996099217723202e-06,
+      "loss": 0.4881,
+      "step": 10759
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1268658529831788,
+      "learning_rate": 1.3983281574170271e-06,
+      "loss": 0.4501,
+      "step": 10760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2337646423986006,
+      "learning_rate": 1.3970469361296557e-06,
+      "loss": 0.4722,
+      "step": 10761
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.251927130037995,
+      "learning_rate": 1.3957662579910969e-06,
+      "loss": 0.4932,
+      "step": 10762
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2588248118317495,
+      "learning_rate": 1.394486123082206e-06,
+      "loss": 0.4983,
+      "step": 10763
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3571947964431077,
+      "learning_rate": 1.3932065314838071e-06,
+      "loss": 0.5309,
+      "step": 10764
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1875720253233086,
+      "learning_rate": 1.3919274832766838e-06,
+      "loss": 0.4871,
+      "step": 10765
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2049814244180048,
+      "learning_rate": 1.390648978541589e-06,
+      "loss": 0.489,
+      "step": 10766
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1682337329563806,
+      "learning_rate": 1.3893710173592457e-06,
+      "loss": 0.5027,
+      "step": 10767
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2058632146528905,
+      "learning_rate": 1.3880935998103317e-06,
+      "loss": 0.5019,
+      "step": 10768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1302106244932704,
+      "learning_rate": 1.386816725975504e-06,
+      "loss": 0.4566,
+      "step": 10769
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2500268456437313,
+      "learning_rate": 1.385540395935372e-06,
+      "loss": 0.5164,
+      "step": 10770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2638334144509422,
+      "learning_rate": 1.3842646097705193e-06,
+      "loss": 0.4984,
+      "step": 10771
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2933552328611988,
+      "learning_rate": 1.3829893675614924e-06,
+      "loss": 0.4955,
+      "step": 10772
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0794999573898838,
+      "learning_rate": 1.381714669388805e-06,
+      "loss": 0.471,
+      "step": 10773
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0502251497201327,
+      "learning_rate": 1.380440515332938e-06,
+      "loss": 0.4248,
+      "step": 10774
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2664682852007916,
+      "learning_rate": 1.3791669054743295e-06,
+      "loss": 0.5356,
+      "step": 10775
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2468282991999549,
+      "learning_rate": 1.3778938398933927e-06,
+      "loss": 0.5251,
+      "step": 10776
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1621070156554307,
+      "learning_rate": 1.3766213186705036e-06,
+      "loss": 0.4653,
+      "step": 10777
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.294201496602907,
+      "learning_rate": 1.3753493418859987e-06,
+      "loss": 0.5818,
+      "step": 10778
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2030049549010255,
+      "learning_rate": 1.3740779096201883e-06,
+      "loss": 0.5253,
+      "step": 10779
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2743483953986934,
+      "learning_rate": 1.372807021953345e-06,
+      "loss": 0.5098,
+      "step": 10780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2291839188911362,
+      "learning_rate": 1.3715366789657025e-06,
+      "loss": 0.5083,
+      "step": 10781
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2275395480620062,
+      "learning_rate": 1.3702668807374664e-06,
+      "loss": 0.5127,
+      "step": 10782
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.235526212024401,
+      "learning_rate": 1.368997627348806e-06,
+      "loss": 0.5532,
+      "step": 10783
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.149829151067626,
+      "learning_rate": 1.367728918879856e-06,
+      "loss": 0.4838,
+      "step": 10784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2707991621499186,
+      "learning_rate": 1.3664607554107178e-06,
+      "loss": 0.5214,
+      "step": 10785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1669629719800076,
+      "learning_rate": 1.3651931370214533e-06,
+      "loss": 0.4473,
+      "step": 10786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1423184613295985,
+      "learning_rate": 1.3639260637920971e-06,
+      "loss": 0.4627,
+      "step": 10787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1849397364716565,
+      "learning_rate": 1.3626595358026429e-06,
+      "loss": 0.4989,
+      "step": 10788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3661531295707572,
+      "learning_rate": 1.3613935531330558e-06,
+      "loss": 0.5729,
+      "step": 10789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1020821502236913,
+      "learning_rate": 1.360128115863265e-06,
+      "loss": 0.4742,
+      "step": 10790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.198002599315339,
+      "learning_rate": 1.3588632240731591e-06,
+      "loss": 0.531,
+      "step": 10791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2487884849292366,
+      "learning_rate": 1.3575988778426008e-06,
+      "loss": 0.4717,
+      "step": 10792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2853957303823997,
+      "learning_rate": 1.356335077251415e-06,
+      "loss": 0.5587,
+      "step": 10793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.16087185659976,
+      "learning_rate": 1.35507182237939e-06,
+      "loss": 0.4719,
+      "step": 10794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0618740369625115,
+      "learning_rate": 1.3538091133062858e-06,
+      "loss": 0.4866,
+      "step": 10795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4409372450580755,
+      "learning_rate": 1.3525469501118183e-06,
+      "loss": 0.5358,
+      "step": 10796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2967474541507673,
+      "learning_rate": 1.3512853328756792e-06,
+      "loss": 0.5178,
+      "step": 10797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1611695151467718,
+      "learning_rate": 1.350024261677516e-06,
+      "loss": 0.4664,
+      "step": 10798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1657886380037057,
+      "learning_rate": 1.3487637365969497e-06,
+      "loss": 0.5048,
+      "step": 10799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3451269768405352,
+      "learning_rate": 1.347503757713564e-06,
+      "loss": 0.5801,
+      "step": 10800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1820919467548117,
+      "learning_rate": 1.3462443251069069e-06,
+      "loss": 0.4986,
+      "step": 10801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.175646904148628,
+      "learning_rate": 1.344985438856492e-06,
+      "loss": 0.4639,
+      "step": 10802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2158919347785935,
+      "learning_rate": 1.3437270990418005e-06,
+      "loss": 0.4836,
+      "step": 10803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1412074548451747,
+      "learning_rate": 1.3424693057422778e-06,
+      "loss": 0.4691,
+      "step": 10804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2210774328468081,
+      "learning_rate": 1.3412120590373368e-06,
+      "loss": 0.4906,
+      "step": 10805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3015930575187398,
+      "learning_rate": 1.3399553590063496e-06,
+      "loss": 0.4965,
+      "step": 10806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2948082691375216,
+      "learning_rate": 1.338699205728663e-06,
+      "loss": 0.5724,
+      "step": 10807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1802022455142054,
+      "learning_rate": 1.3374435992835798e-06,
+      "loss": 0.4926,
+      "step": 10808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1665733958608424,
+      "learning_rate": 1.3361885397503749e-06,
+      "loss": 0.5317,
+      "step": 10809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2625637000603989,
+      "learning_rate": 1.334934027208289e-06,
+      "loss": 0.5471,
+      "step": 10810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3277707244485306,
+      "learning_rate": 1.333680061736522e-06,
+      "loss": 0.5281,
+      "step": 10811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2786571970924778,
+      "learning_rate": 1.3324266434142452e-06,
+      "loss": 0.5368,
+      "step": 10812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3105343677361685,
+      "learning_rate": 1.331173772320593e-06,
+      "loss": 0.5353,
+      "step": 10813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.213867531221668,
+      "learning_rate": 1.3299214485346657e-06,
+      "loss": 0.536,
+      "step": 10814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1527844636919835,
+      "learning_rate": 1.3286696721355308e-06,
+      "loss": 0.5245,
+      "step": 10815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3641731820313083,
+      "learning_rate": 1.3274184432022163e-06,
+      "loss": 0.5001,
+      "step": 10816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1386973670998044,
+      "learning_rate": 1.3261677618137225e-06,
+      "loss": 0.5058,
+      "step": 10817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.208461765610928,
+      "learning_rate": 1.3249176280490062e-06,
+      "loss": 0.5306,
+      "step": 10818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.19888057351629,
+      "learning_rate": 1.3236680419869974e-06,
+      "loss": 0.5333,
+      "step": 10819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.17742789410155,
+      "learning_rate": 1.322419003706592e-06,
+      "loss": 0.5125,
+      "step": 10820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4860746591124063,
+      "learning_rate": 1.3211705132866425e-06,
+      "loss": 0.5186,
+      "step": 10821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2032560301930184,
+      "learning_rate": 1.319922570805976e-06,
+      "loss": 0.4705,
+      "step": 10822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0833945379329721,
+      "learning_rate": 1.3186751763433803e-06,
+      "loss": 0.4589,
+      "step": 10823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1397373455706485,
+      "learning_rate": 1.3174283299776103e-06,
+      "loss": 0.4907,
+      "step": 10824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2937090272911513,
+      "learning_rate": 1.3161820317873886e-06,
+      "loss": 0.5254,
+      "step": 10825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2711673933288596,
+      "learning_rate": 1.3149362818513955e-06,
+      "loss": 0.4953,
+      "step": 10826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2915378731946339,
+      "learning_rate": 1.3136910802482862e-06,
+      "loss": 0.5293,
+      "step": 10827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.238818656720174,
+      "learning_rate": 1.3124464270566727e-06,
+      "loss": 0.4607,
+      "step": 10828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2065017180606452,
+      "learning_rate": 1.3112023223551374e-06,
+      "loss": 0.5077,
+      "step": 10829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1675112029034365,
+      "learning_rate": 1.3099587662222302e-06,
+      "loss": 0.4655,
+      "step": 10830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1661638073077676,
+      "learning_rate": 1.3087157587364596e-06,
+      "loss": 0.4585,
+      "step": 10831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.175343530230575,
+      "learning_rate": 1.3074732999763029e-06,
+      "loss": 0.5753,
+      "step": 10832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1335883279114707,
+      "learning_rate": 1.306231390020205e-06,
+      "loss": 0.4821,
+      "step": 10833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2057703833679223,
+      "learning_rate": 1.3049900289465733e-06,
+      "loss": 0.5463,
+      "step": 10834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0682319411687036,
+      "learning_rate": 1.303749216833784e-06,
+      "loss": 0.456,
+      "step": 10835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.247203846640096,
+      "learning_rate": 1.30250895376017e-06,
+      "loss": 0.5179,
+      "step": 10836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1588513777910028,
+      "learning_rate": 1.3012692398040416e-06,
+      "loss": 0.4801,
+      "step": 10837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2390458787762963,
+      "learning_rate": 1.3000300750436645e-06,
+      "loss": 0.573,
+      "step": 10838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1628762671955695,
+      "learning_rate": 1.2987914595572738e-06,
+      "loss": 0.5235,
+      "step": 10839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2102574069137875,
+      "learning_rate": 1.297553393423071e-06,
+      "loss": 0.5248,
+      "step": 10840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1891242513775562,
+      "learning_rate": 1.296315876719223e-06,
+      "loss": 0.499,
+      "step": 10841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3099195909210037,
+      "learning_rate": 1.295078909523857e-06,
+      "loss": 0.5159,
+      "step": 10842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1284373012258375,
+      "learning_rate": 1.2938424919150705e-06,
+      "loss": 0.5088,
+      "step": 10843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.188858610259817,
+      "learning_rate": 1.2926066239709256e-06,
+      "loss": 0.5223,
+      "step": 10844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.190536481715684,
+      "learning_rate": 1.2913713057694498e-06,
+      "loss": 0.5015,
+      "step": 10845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2773474363687967,
+      "learning_rate": 1.2901365373886331e-06,
+      "loss": 0.5389,
+      "step": 10846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2536104988387609,
+      "learning_rate": 1.2889023189064364e-06,
+      "loss": 0.5022,
+      "step": 10847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.208849429486005,
+      "learning_rate": 1.2876686504007764e-06,
+      "loss": 0.4989,
+      "step": 10848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1443225032280204,
+      "learning_rate": 1.2864355319495448e-06,
+      "loss": 0.4371,
+      "step": 10849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1924101055498921,
+      "learning_rate": 1.285202963630594e-06,
+      "loss": 0.5226,
+      "step": 10850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2944129645776967,
+      "learning_rate": 1.2839709455217453e-06,
+      "loss": 0.5451,
+      "step": 10851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2238177063148723,
+      "learning_rate": 1.2827394777007774e-06,
+      "loss": 0.5202,
+      "step": 10852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1125773027972965,
+      "learning_rate": 1.2815085602454401e-06,
+      "loss": 0.4273,
+      "step": 10853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2329779819679016,
+      "learning_rate": 1.2802781932334495e-06,
+      "loss": 0.5011,
+      "step": 10854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.157375715598917,
+      "learning_rate": 1.2790483767424878e-06,
+      "loss": 0.4952,
+      "step": 10855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.310867293140858,
+      "learning_rate": 1.2778191108501925e-06,
+      "loss": 0.5292,
+      "step": 10856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.193135142135876,
+      "learning_rate": 1.2765903956341807e-06,
+      "loss": 0.4734,
+      "step": 10857
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.15281770948524,
+      "learning_rate": 1.2753622311720203e-06,
+      "loss": 0.4928,
+      "step": 10858
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2165344837097434,
+      "learning_rate": 1.2741346175412571e-06,
+      "loss": 0.4891,
+      "step": 10859
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3436418977768463,
+      "learning_rate": 1.2729075548193947e-06,
+      "loss": 0.6186,
+      "step": 10860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1532568982784446,
+      "learning_rate": 1.2716810430839056e-06,
+      "loss": 0.5101,
+      "step": 10861
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2762731329215957,
+      "learning_rate": 1.270455082412223e-06,
+      "loss": 0.5155,
+      "step": 10862
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1895692264413007,
+      "learning_rate": 1.2692296728817487e-06,
+      "loss": 0.5011,
+      "step": 10863
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2093423489049155,
+      "learning_rate": 1.268004814569851e-06,
+      "loss": 0.5906,
+      "step": 10864
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1630280269679925,
+      "learning_rate": 1.266780507553863e-06,
+      "loss": 0.4933,
+      "step": 10865
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.017427107428367,
+      "learning_rate": 1.2655567519110756e-06,
+      "loss": 0.4223,
+      "step": 10866
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1516943184550408,
+      "learning_rate": 1.2643335477187567e-06,
+      "loss": 0.5277,
+      "step": 10867
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1864273346336496,
+      "learning_rate": 1.2631108950541303e-06,
+      "loss": 0.4524,
+      "step": 10868
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1194909630905223,
+      "learning_rate": 1.261888793994388e-06,
+      "loss": 0.5079,
+      "step": 10869
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1862825376753299,
+      "learning_rate": 1.260667244616689e-06,
+      "loss": 0.4726,
+      "step": 10870
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.122631546842427,
+      "learning_rate": 1.2594462469981582e-06,
+      "loss": 0.4809,
+      "step": 10871
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.098077659216121,
+      "learning_rate": 1.2582258012158799e-06,
+      "loss": 0.4574,
+      "step": 10872
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1962322112607366,
+      "learning_rate": 1.2570059073469076e-06,
+      "loss": 0.4667,
+      "step": 10873
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2463609175838648,
+      "learning_rate": 1.25578656546826e-06,
+      "loss": 0.5413,
+      "step": 10874
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2412484896401919,
+      "learning_rate": 1.254567775656923e-06,
+      "loss": 0.4892,
+      "step": 10875
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1532129145658276,
+      "learning_rate": 1.2533495379898407e-06,
+      "loss": 0.4851,
+      "step": 10876
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1250474707866498,
+      "learning_rate": 1.252131852543932e-06,
+      "loss": 0.471,
+      "step": 10877
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.241355425380615,
+      "learning_rate": 1.250914719396069e-06,
+      "loss": 0.5385,
+      "step": 10878
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1449373846132629,
+      "learning_rate": 1.2496981386231e-06,
+      "loss": 0.472,
+      "step": 10879
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2023882901652603,
+      "learning_rate": 1.2484821103018329e-06,
+      "loss": 0.5339,
+      "step": 10880
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.168978914496552,
+      "learning_rate": 1.2472666345090435e-06,
+      "loss": 0.4872,
+      "step": 10881
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.106973278005702,
+      "learning_rate": 1.2460517113214688e-06,
+      "loss": 0.4774,
+      "step": 10882
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0929780143034311,
+      "learning_rate": 1.2448373408158133e-06,
+      "loss": 0.4846,
+      "step": 10883
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2446006511296048,
+      "learning_rate": 1.2436235230687466e-06,
+      "loss": 0.5218,
+      "step": 10884
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.138021712080953,
+      "learning_rate": 1.2424102581569064e-06,
+      "loss": 0.4933,
+      "step": 10885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1251463794844778,
+      "learning_rate": 1.2411975461568881e-06,
+      "loss": 0.493,
+      "step": 10886
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2287749242728667,
+      "learning_rate": 1.2399853871452605e-06,
+      "loss": 0.514,
+      "step": 10887
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.206398955719468,
+      "learning_rate": 1.2387737811985479e-06,
+      "loss": 0.4976,
+      "step": 10888
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.208941580611243,
+      "learning_rate": 1.23756272839325e-06,
+      "loss": 0.5029,
+      "step": 10889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.255789984748273,
+      "learning_rate": 1.2363522288058238e-06,
+      "loss": 0.5212,
+      "step": 10890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.258270038292831,
+      "learning_rate": 1.2351422825126969e-06,
+      "loss": 0.4995,
+      "step": 10891
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2292192200024226,
+      "learning_rate": 1.2339328895902603e-06,
+      "loss": 0.5207,
+      "step": 10892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1894136870161656,
+      "learning_rate": 1.2327240501148651e-06,
+      "loss": 0.4845,
+      "step": 10893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2650543621881953,
+      "learning_rate": 1.2315157641628338e-06,
+      "loss": 0.5372,
+      "step": 10894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.198015684375401,
+      "learning_rate": 1.2303080318104533e-06,
+      "loss": 0.4852,
+      "step": 10895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1911042878256393,
+      "learning_rate": 1.229100853133971e-06,
+      "loss": 0.4934,
+      "step": 10896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2255815118352613,
+      "learning_rate": 1.2278942282096063e-06,
+      "loss": 0.4829,
+      "step": 10897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2046207196716854,
+      "learning_rate": 1.2266881571135337e-06,
+      "loss": 0.5593,
+      "step": 10898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1641985410524214,
+      "learning_rate": 1.2254826399219032e-06,
+      "loss": 0.473,
+      "step": 10899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.309079527007699,
+      "learning_rate": 1.2242776767108233e-06,
+      "loss": 0.5477,
+      "step": 10900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2956451191423466,
+      "learning_rate": 1.2230732675563705e-06,
+      "loss": 0.5157,
+      "step": 10901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.318197147136756,
+      "learning_rate": 1.2218694125345887e-06,
+      "loss": 0.5097,
+      "step": 10902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2899736363282295,
+      "learning_rate": 1.2206661117214768e-06,
+      "loss": 0.5074,
+      "step": 10903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1875261002734072,
+      "learning_rate": 1.2194633651930089e-06,
+      "loss": 0.447,
+      "step": 10904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1686532556812843,
+      "learning_rate": 1.2182611730251225e-06,
+      "loss": 0.4576,
+      "step": 10905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1182941211206503,
+      "learning_rate": 1.2170595352937142e-06,
+      "loss": 0.4488,
+      "step": 10906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.26196277254576,
+      "learning_rate": 1.215858452074653e-06,
+      "loss": 0.4595,
+      "step": 10907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.26661469181406,
+      "learning_rate": 1.214657923443766e-06,
+      "loss": 0.5371,
+      "step": 10908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1408964382818523,
+      "learning_rate": 1.2134579494768507e-06,
+      "loss": 0.4682,
+      "step": 10909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1716665972731792,
+      "learning_rate": 1.2122585302496682e-06,
+      "loss": 0.4829,
+      "step": 10910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2747453397855322,
+      "learning_rate": 1.2110596658379426e-06,
+      "loss": 0.5306,
+      "step": 10911
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2727065065318452,
+      "learning_rate": 1.2098613563173678e-06,
+      "loss": 0.5048,
+      "step": 10912
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0808364385963682,
+      "learning_rate": 1.2086636017635955e-06,
+      "loss": 0.4559,
+      "step": 10913
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1184747389258711,
+      "learning_rate": 1.2074664022522464e-06,
+      "loss": 0.4905,
+      "step": 10914
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1918389707099555,
+      "learning_rate": 1.2062697578589089e-06,
+      "loss": 0.4776,
+      "step": 10915
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1282726476531821,
+      "learning_rate": 1.2050736686591292e-06,
+      "loss": 0.4776,
+      "step": 10916
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1985809916668981,
+      "learning_rate": 1.2038781347284266e-06,
+      "loss": 0.4855,
+      "step": 10917
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2888363726557173,
+      "learning_rate": 1.202683156142278e-06,
+      "loss": 0.5117,
+      "step": 10918
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3553973352269555,
+      "learning_rate": 1.2014887329761293e-06,
+      "loss": 0.4993,
+      "step": 10919
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2015193777692599,
+      "learning_rate": 1.2002948653053915e-06,
+      "loss": 0.5126,
+      "step": 10920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2364997925378338,
+      "learning_rate": 1.1991015532054395e-06,
+      "loss": 0.5,
+      "step": 10921
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1539568973240086,
+      "learning_rate": 1.1979087967516146e-06,
+      "loss": 0.4963,
+      "step": 10922
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2573862241992435,
+      "learning_rate": 1.1967165960192185e-06,
+      "loss": 0.5441,
+      "step": 10923
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2326866385415258,
+      "learning_rate": 1.1955249510835232e-06,
+      "loss": 0.4765,
+      "step": 10924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2356714128606008,
+      "learning_rate": 1.1943338620197642e-06,
+      "loss": 0.4759,
+      "step": 10925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3008147455891728,
+      "learning_rate": 1.193143328903138e-06,
+      "loss": 0.484,
+      "step": 10926
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2964257070996899,
+      "learning_rate": 1.1919533518088121e-06,
+      "loss": 0.5137,
+      "step": 10927
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.148413054536943,
+      "learning_rate": 1.1907639308119134e-06,
+      "loss": 0.4726,
+      "step": 10928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2314015561466816,
+      "learning_rate": 1.1895750659875372e-06,
+      "loss": 0.5329,
+      "step": 10929
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2561606228287285,
+      "learning_rate": 1.1883867574107433e-06,
+      "loss": 0.5213,
+      "step": 10930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.160572632330765,
+      "learning_rate": 1.1871990051565551e-06,
+      "loss": 0.4701,
+      "step": 10931
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1548002794497303,
+      "learning_rate": 1.186011809299964e-06,
+      "loss": 0.535,
+      "step": 10932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.091084146994098,
+      "learning_rate": 1.1848251699159185e-06,
+      "loss": 0.437,
+      "step": 10933
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3319150364987997,
+      "learning_rate": 1.1836390870793414e-06,
+      "loss": 0.5169,
+      "step": 10934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.301156158947863,
+      "learning_rate": 1.1824535608651177e-06,
+      "loss": 0.5143,
+      "step": 10935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2165011173561442,
+      "learning_rate": 1.1812685913480904e-06,
+      "loss": 0.5389,
+      "step": 10936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2067531524579094,
+      "learning_rate": 1.180084178603077e-06,
+      "loss": 0.4974,
+      "step": 10937
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1960559103952126,
+      "learning_rate": 1.1789003227048533e-06,
+      "loss": 0.4589,
+      "step": 10938
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2521339797674484,
+      "learning_rate": 1.1777170237281633e-06,
+      "loss": 0.5686,
+      "step": 10939
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2981094092735477,
+      "learning_rate": 1.1765342817477133e-06,
+      "loss": 0.5404,
+      "step": 10940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.216255912727184,
+      "learning_rate": 1.1753520968381782e-06,
+      "loss": 0.4981,
+      "step": 10941
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1948853770552794,
+      "learning_rate": 1.1741704690741961e-06,
+      "loss": 0.5024,
+      "step": 10942
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2560423722135725,
+      "learning_rate": 1.1729893985303653e-06,
+      "loss": 0.537,
+      "step": 10943
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2590150948332368,
+      "learning_rate": 1.171808885281256e-06,
+      "loss": 0.501,
+      "step": 10944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.240003236874078,
+      "learning_rate": 1.1706289294014005e-06,
+      "loss": 0.5355,
+      "step": 10945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.139649012809811,
+      "learning_rate": 1.169449530965292e-06,
+      "loss": 0.4487,
+      "step": 10946
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2459014457191502,
+      "learning_rate": 1.168270690047395e-06,
+      "loss": 0.5409,
+      "step": 10947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.178618799636493,
+      "learning_rate": 1.1670924067221367e-06,
+      "loss": 0.499,
+      "step": 10948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.146812431021523,
+      "learning_rate": 1.1659146810639043e-06,
+      "loss": 0.5009,
+      "step": 10949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2949706193544122,
+      "learning_rate": 1.1647375131470562e-06,
+      "loss": 0.4855,
+      "step": 10950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2314065901425824,
+      "learning_rate": 1.1635609030459127e-06,
+      "loss": 0.532,
+      "step": 10951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1257868769953112,
+      "learning_rate": 1.1623848508347603e-06,
+      "loss": 0.4624,
+      "step": 10952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.134018880468286,
+      "learning_rate": 1.1612093565878502e-06,
+      "loss": 0.5205,
+      "step": 10953
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3056232956746885,
+      "learning_rate": 1.1600344203793922e-06,
+      "loss": 0.585,
+      "step": 10954
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0941822287873288,
+      "learning_rate": 1.1588600422835728e-06,
+      "loss": 0.4613,
+      "step": 10955
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3033279370702722,
+      "learning_rate": 1.157686222374531e-06,
+      "loss": 0.5306,
+      "step": 10956
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1631717730284945,
+      "learning_rate": 1.156512960726378e-06,
+      "loss": 0.5017,
+      "step": 10957
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1042728492975822,
+      "learning_rate": 1.15534025741319e-06,
+      "loss": 0.5094,
+      "step": 10958
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1033958167470623,
+      "learning_rate": 1.1541681125090031e-06,
+      "loss": 0.4616,
+      "step": 10959
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2065354597474753,
+      "learning_rate": 1.1529965260878207e-06,
+      "loss": 0.5443,
+      "step": 10960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2799111791430533,
+      "learning_rate": 1.1518254982236121e-06,
+      "loss": 0.512,
+      "step": 10961
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2412995337584116,
+      "learning_rate": 1.1506550289903107e-06,
+      "loss": 0.5096,
+      "step": 10962
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1974353763289312,
+      "learning_rate": 1.149485118461816e-06,
+      "loss": 0.4898,
+      "step": 10963
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1782457750173052,
+      "learning_rate": 1.148315766711986e-06,
+      "loss": 0.4912,
+      "step": 10964
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2835463437911176,
+      "learning_rate": 1.1471469738146534e-06,
+      "loss": 0.5307,
+      "step": 10965
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1431393541005277,
+      "learning_rate": 1.145978739843604e-06,
+      "loss": 0.4886,
+      "step": 10966
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3437348741966413,
+      "learning_rate": 1.1448110648725974e-06,
+      "loss": 0.4795,
+      "step": 10967
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1581387429572234,
+      "learning_rate": 1.1436439489753581e-06,
+      "loss": 0.4731,
+      "step": 10968
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1961365894541909,
+      "learning_rate": 1.1424773922255662e-06,
+      "loss": 0.488,
+      "step": 10969
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3464965695813065,
+      "learning_rate": 1.1413113946968756e-06,
+      "loss": 0.5224,
+      "step": 10970
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2091670229591636,
+      "learning_rate": 1.1401459564629013e-06,
+      "loss": 0.5662,
+      "step": 10971
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.235892316954457,
+      "learning_rate": 1.1389810775972244e-06,
+      "loss": 0.5629,
+      "step": 10972
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2497525447049591,
+      "learning_rate": 1.1378167581733901e-06,
+      "loss": 0.5027,
+      "step": 10973
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2061069245551066,
+      "learning_rate": 1.1366529982649043e-06,
+      "loss": 0.5051,
+      "step": 10974
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2388547898823448,
+      "learning_rate": 1.1354897979452472e-06,
+      "loss": 0.5195,
+      "step": 10975
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1407296642202176,
+      "learning_rate": 1.13432715728785e-06,
+      "loss": 0.4769,
+      "step": 10976
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.212526570845022,
+      "learning_rate": 1.1331650763661217e-06,
+      "loss": 0.5651,
+      "step": 10977
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2109669343385894,
+      "learning_rate": 1.1320035552534304e-06,
+      "loss": 0.5194,
+      "step": 10978
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.152942876725815,
+      "learning_rate": 1.130842594023106e-06,
+      "loss": 0.452,
+      "step": 10979
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0769316108334046,
+      "learning_rate": 1.129682192748447e-06,
+      "loss": 0.5106,
+      "step": 10980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2568610721615041,
+      "learning_rate": 1.1285223515027155e-06,
+      "loss": 0.5194,
+      "step": 10981
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2569610841655474,
+      "learning_rate": 1.12736307035914e-06,
+      "loss": 0.5669,
+      "step": 10982
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2924073720682197,
+      "learning_rate": 1.1262043493909113e-06,
+      "loss": 0.5731,
+      "step": 10983
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1220376430765475,
+      "learning_rate": 1.125046188671184e-06,
+      "loss": 0.5025,
+      "step": 10984
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.255659689682778,
+      "learning_rate": 1.123888588273081e-06,
+      "loss": 0.5235,
+      "step": 10985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2200607196362419,
+      "learning_rate": 1.1227315482696844e-06,
+      "loss": 0.4696,
+      "step": 10986
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2064723724003361,
+      "learning_rate": 1.1215750687340455e-06,
+      "loss": 0.4811,
+      "step": 10987
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1083721947609209,
+      "learning_rate": 1.1204191497391815e-06,
+      "loss": 0.4939,
+      "step": 10988
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2110291477100763,
+      "learning_rate": 1.119263791358066e-06,
+      "loss": 0.5515,
+      "step": 10989
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2576024608694756,
+      "learning_rate": 1.118108993663647e-06,
+      "loss": 0.546,
+      "step": 10990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1645289772919507,
+      "learning_rate": 1.1169547567288319e-06,
+      "loss": 0.4856,
+      "step": 10991
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0279983349723543,
+      "learning_rate": 1.115801080626493e-06,
+      "loss": 0.4487,
+      "step": 10992
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.210358265810039,
+      "learning_rate": 1.1146479654294706e-06,
+      "loss": 0.5128,
+      "step": 10993
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2835304620781385,
+      "learning_rate": 1.1134954112105645e-06,
+      "loss": 0.5464,
+      "step": 10994
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2349588968796155,
+      "learning_rate": 1.1123434180425396e-06,
+      "loss": 0.494,
+      "step": 10995
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2615550971410479,
+      "learning_rate": 1.1111919859981291e-06,
+      "loss": 0.5084,
+      "step": 10996
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3699009589433255,
+      "learning_rate": 1.1100411151500279e-06,
+      "loss": 0.5049,
+      "step": 10997
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1878724267099376,
+      "learning_rate": 1.1088908055709003e-06,
+      "loss": 0.564,
+      "step": 10998
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2212558807106515,
+      "learning_rate": 1.1077410573333659e-06,
+      "loss": 0.5244,
+      "step": 10999
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1960268067924829,
+      "learning_rate": 1.1065918705100164e-06,
+      "loss": 0.4708,
+      "step": 11000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.236982464702292,
+      "learning_rate": 1.1054432451734053e-06,
+      "loss": 0.5139,
+      "step": 11001
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2934956471031154,
+      "learning_rate": 1.1042951813960535e-06,
+      "loss": 0.499,
+      "step": 11002
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.185914738251953,
+      "learning_rate": 1.1031476792504436e-06,
+      "loss": 0.4702,
+      "step": 11003
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3610791015537416,
+      "learning_rate": 1.1020007388090227e-06,
+      "loss": 0.5566,
+      "step": 11004
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1509418300409455,
+      "learning_rate": 1.1008543601442012e-06,
+      "loss": 0.4898,
+      "step": 11005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0690610488752403,
+      "learning_rate": 1.0997085433283572e-06,
+      "loss": 0.4547,
+      "step": 11006
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.172408224227052,
+      "learning_rate": 1.0985632884338327e-06,
+      "loss": 0.5034,
+      "step": 11007
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.152837149815908,
+      "learning_rate": 1.097418595532933e-06,
+      "loss": 0.5033,
+      "step": 11008
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.18503761980242,
+      "learning_rate": 1.0962744646979296e-06,
+      "loss": 0.5057,
+      "step": 11009
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3533746114708962,
+      "learning_rate": 1.0951308960010554e-06,
+      "loss": 0.5416,
+      "step": 11010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1421855545549224,
+      "learning_rate": 1.0939878895145107e-06,
+      "loss": 0.4833,
+      "step": 11011
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2718850250223959,
+      "learning_rate": 1.0928454453104597e-06,
+      "loss": 0.5475,
+      "step": 11012
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1922615858184182,
+      "learning_rate": 1.0917035634610328e-06,
+      "loss": 0.5016,
+      "step": 11013
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.230251571623262,
+      "learning_rate": 1.09056224403832e-06,
+      "loss": 0.5077,
+      "step": 11014
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.129394795128725,
+      "learning_rate": 1.0894214871143783e-06,
+      "loss": 0.4975,
+      "step": 11015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1328901066998234,
+      "learning_rate": 1.0882812927612297e-06,
+      "loss": 0.4733,
+      "step": 11016
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2175706390160521,
+      "learning_rate": 1.0871416610508622e-06,
+      "loss": 0.5217,
+      "step": 11017
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2137954457473987,
+      "learning_rate": 1.0860025920552252e-06,
+      "loss": 0.4689,
+      "step": 11018
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2838029780316345,
+      "learning_rate": 1.0848640858462378e-06,
+      "loss": 0.5279,
+      "step": 11019
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1396354668017237,
+      "learning_rate": 1.0837261424957735e-06,
+      "loss": 0.4426,
+      "step": 11020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2505302258315805,
+      "learning_rate": 1.0825887620756804e-06,
+      "loss": 0.5382,
+      "step": 11021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.226496360599236,
+      "learning_rate": 1.0814519446577665e-06,
+      "loss": 0.5071,
+      "step": 11022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2339909112895453,
+      "learning_rate": 1.0803156903138069e-06,
+      "loss": 0.5321,
+      "step": 11023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1292351372335896,
+      "learning_rate": 1.0791799991155371e-06,
+      "loss": 0.4608,
+      "step": 11024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.178304910456379,
+      "learning_rate": 1.078044871134658e-06,
+      "loss": 0.4722,
+      "step": 11025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0947160541666423,
+      "learning_rate": 1.0769103064428366e-06,
+      "loss": 0.4624,
+      "step": 11026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1775900276254219,
+      "learning_rate": 1.0757763051117055e-06,
+      "loss": 0.5176,
+      "step": 11027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2413996471231832,
+      "learning_rate": 1.0746428672128583e-06,
+      "loss": 0.4981,
+      "step": 11028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1791926476589687,
+      "learning_rate": 1.0735099928178584e-06,
+      "loss": 0.4937,
+      "step": 11029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4016056083930608,
+      "learning_rate": 1.0723776819982257e-06,
+      "loss": 0.5165,
+      "step": 11030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3500335071078888,
+      "learning_rate": 1.0712459348254488e-06,
+      "loss": 0.5668,
+      "step": 11031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.297593848228336,
+      "learning_rate": 1.070114751370984e-06,
+      "loss": 0.5351,
+      "step": 11032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2029376193684853,
+      "learning_rate": 1.0689841317062478e-06,
+      "loss": 0.5146,
+      "step": 11033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2583092603001815,
+      "learning_rate": 1.0678540759026224e-06,
+      "loss": 0.5115,
+      "step": 11034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1310975346117904,
+      "learning_rate": 1.06672458403145e-06,
+      "loss": 0.4644,
+      "step": 11035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.155860061877357,
+      "learning_rate": 1.0655956561640456e-06,
+      "loss": 0.4531,
+      "step": 11036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2350248243948958,
+      "learning_rate": 1.0644672923716815e-06,
+      "loss": 0.5187,
+      "step": 11037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1229920481691575,
+      "learning_rate": 1.063339492725599e-06,
+      "loss": 0.4594,
+      "step": 11038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2306060714389633,
+      "learning_rate": 1.062212257297004e-06,
+      "loss": 0.5002,
+      "step": 11039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1198946291879468,
+      "learning_rate": 1.061085586157059e-06,
+      "loss": 0.4899,
+      "step": 11040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.106094345090458,
+      "learning_rate": 1.0599594793769007e-06,
+      "loss": 0.4546,
+      "step": 11041
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0821773829745938,
+      "learning_rate": 1.0588339370276246e-06,
+      "loss": 0.4569,
+      "step": 11042
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1525582324064279,
+      "learning_rate": 1.0577089591802946e-06,
+      "loss": 0.5114,
+      "step": 11043
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2049383889272243,
+      "learning_rate": 1.0565845459059343e-06,
+      "loss": 0.5388,
+      "step": 11044
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2112109460036242,
+      "learning_rate": 1.055460697275531e-06,
+      "loss": 0.4769,
+      "step": 11045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2140897514593778,
+      "learning_rate": 1.0543374133600414e-06,
+      "loss": 0.5327,
+      "step": 11046
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2406017810560912,
+      "learning_rate": 1.0532146942303856e-06,
+      "loss": 0.4614,
+      "step": 11047
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1999057454604056,
+      "learning_rate": 1.0520925399574445e-06,
+      "loss": 0.5037,
+      "step": 11048
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3304978659238906,
+      "learning_rate": 1.050970950612069e-06,
+      "loss": 0.5186,
+      "step": 11049
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.313559785496592,
+      "learning_rate": 1.0498499262650664e-06,
+      "loss": 0.5166,
+      "step": 11050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4039201721961587,
+      "learning_rate": 1.048729466987214e-06,
+      "loss": 0.5652,
+      "step": 11051
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1895264350836758,
+      "learning_rate": 1.047609572849253e-06,
+      "loss": 0.4987,
+      "step": 11052
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2395595366600698,
+      "learning_rate": 1.0464902439218905e-06,
+      "loss": 0.4588,
+      "step": 11053
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1489206452585214,
+      "learning_rate": 1.0453714802757908e-06,
+      "loss": 0.4597,
+      "step": 11054
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.131600882978287,
+      "learning_rate": 1.0442532819815908e-06,
+      "loss": 0.4819,
+      "step": 11055
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.239627960213853,
+      "learning_rate": 1.0431356491098854e-06,
+      "loss": 0.5627,
+      "step": 11056
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.228975776893027,
+      "learning_rate": 1.0420185817312377e-06,
+      "loss": 0.5407,
+      "step": 11057
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3174200031525334,
+      "learning_rate": 1.040902079916174e-06,
+      "loss": 0.4887,
+      "step": 11058
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.124350678295888,
+      "learning_rate": 1.0397861437351842e-06,
+      "loss": 0.5002,
+      "step": 11059
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4144751858090445,
+      "learning_rate": 1.0386707732587265e-06,
+      "loss": 0.5842,
+      "step": 11060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2034571300679713,
+      "learning_rate": 1.0375559685572145e-06,
+      "loss": 0.4942,
+      "step": 11061
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2950542031362913,
+      "learning_rate": 1.0364417297010354e-06,
+      "loss": 0.5092,
+      "step": 11062
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.300741796496622,
+      "learning_rate": 1.0353280567605373e-06,
+      "loss": 0.4402,
+      "step": 11063
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1856222364426632,
+      "learning_rate": 1.0342149498060284e-06,
+      "loss": 0.4901,
+      "step": 11064
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2007141233505538,
+      "learning_rate": 1.033102408907789e-06,
+      "loss": 0.4733,
+      "step": 11065
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2115407118118497,
+      "learning_rate": 1.0319904341360554e-06,
+      "loss": 0.5184,
+      "step": 11066
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2610145706012588,
+      "learning_rate": 1.0308790255610356e-06,
+      "loss": 0.5298,
+      "step": 11067
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1800056685295375,
+      "learning_rate": 1.0297681832528971e-06,
+      "loss": 0.4939,
+      "step": 11068
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.130417811703316,
+      "learning_rate": 1.0286579072817726e-06,
+      "loss": 0.4671,
+      "step": 11069
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.337200170404787,
+      "learning_rate": 1.0275481977177638e-06,
+      "loss": 0.5938,
+      "step": 11070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1450482655642957,
+      "learning_rate": 1.0264390546309254e-06,
+      "loss": 0.4816,
+      "step": 11071
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.179749493517386,
+      "learning_rate": 1.0253304780912887e-06,
+      "loss": 0.4636,
+      "step": 11072
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1485406219479515,
+      "learning_rate": 1.024222468168843e-06,
+      "loss": 0.4796,
+      "step": 11073
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2519898312524385,
+      "learning_rate": 1.023115024933541e-06,
+      "loss": 0.5411,
+      "step": 11074
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.196609391902362,
+      "learning_rate": 1.0220081484553025e-06,
+      "loss": 0.4884,
+      "step": 11075
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.20362318549424,
+      "learning_rate": 1.0209018388040093e-06,
+      "loss": 0.4696,
+      "step": 11076
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3149237732496482,
+      "learning_rate": 1.019796096049508e-06,
+      "loss": 0.5093,
+      "step": 11077
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1467307766653911,
+      "learning_rate": 1.0186909202616114e-06,
+      "loss": 0.4702,
+      "step": 11078
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1576955368124666,
+      "learning_rate": 1.017586311510095e-06,
+      "loss": 0.4703,
+      "step": 11079
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.216619634807926,
+      "learning_rate": 1.0164822698646992e-06,
+      "loss": 0.5605,
+      "step": 11080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2133083146436283,
+      "learning_rate": 1.0153787953951245e-06,
+      "loss": 0.5514,
+      "step": 11081
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4289492175700447,
+      "learning_rate": 1.0142758881710413e-06,
+      "loss": 0.5101,
+      "step": 11082
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2952194217237716,
+      "learning_rate": 1.013173548262083e-06,
+      "loss": 0.5515,
+      "step": 11083
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1800721911799206,
+      "learning_rate": 1.0120717757378428e-06,
+      "loss": 0.4534,
+      "step": 11084
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3216230670926006,
+      "learning_rate": 1.0109705706678862e-06,
+      "loss": 0.5408,
+      "step": 11085
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2886630277396456,
+      "learning_rate": 1.009869933121731e-06,
+      "loss": 0.4884,
+      "step": 11086
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2858279713444503,
+      "learning_rate": 1.0087698631688713e-06,
+      "loss": 0.5274,
+      "step": 11087
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1717983475093516,
+      "learning_rate": 1.0076703608787575e-06,
+      "loss": 0.5071,
+      "step": 11088
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.311829577382489,
+      "learning_rate": 1.0065714263208092e-06,
+      "loss": 0.5615,
+      "step": 11089
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.238725889261915,
+      "learning_rate": 1.005473059564408e-06,
+      "loss": 0.4752,
+      "step": 11090
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1352940706100734,
+      "learning_rate": 1.004375260678897e-06,
+      "loss": 0.4839,
+      "step": 11091
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2168244041892375,
+      "learning_rate": 1.0032780297335886e-06,
+      "loss": 0.4637,
+      "step": 11092
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1732857414278512,
+      "learning_rate": 1.002181366797753e-06,
+      "loss": 0.5696,
+      "step": 11093
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2030735252565226,
+      "learning_rate": 1.0010852719406306e-06,
+      "loss": 0.5126,
+      "step": 11094
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1032855578928529,
+      "learning_rate": 9.999897452314256e-07,
+      "loss": 0.4557,
+      "step": 11095
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2141840575645526,
+      "learning_rate": 9.988947867392995e-07,
+      "loss": 0.4811,
+      "step": 11096
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1192433574718834,
+      "learning_rate": 9.978003965333849e-07,
+      "loss": 0.4859,
+      "step": 11097
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2169969127932458,
+      "learning_rate": 9.967065746827764e-07,
+      "loss": 0.5249,
+      "step": 11098
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2591108172839338,
+      "learning_rate": 9.956133212565332e-07,
+      "loss": 0.4623,
+      "step": 11099
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2404433667571837,
+      "learning_rate": 9.945206363236804e-07,
+      "loss": 0.5624,
+      "step": 11100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.139843346306126,
+      "learning_rate": 9.93428519953199e-07,
+      "loss": 0.5046,
+      "step": 11101
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0546810573804961,
+      "learning_rate": 9.92336972214044e-07,
+      "loss": 0.4624,
+      "step": 11102
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2546564157704336,
+      "learning_rate": 9.912459931751296e-07,
+      "loss": 0.466,
+      "step": 11103
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2656281082680292,
+      "learning_rate": 9.901555829053333e-07,
+      "loss": 0.5338,
+      "step": 11104
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3375466900100528,
+      "learning_rate": 9.890657414735017e-07,
+      "loss": 0.5899,
+      "step": 11105
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2502162269496075,
+      "learning_rate": 9.879764689484383e-07,
+      "loss": 0.5204,
+      "step": 11106
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.153685432024286,
+      "learning_rate": 9.868877653989161e-07,
+      "loss": 0.4511,
+      "step": 11107
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1333907657451077,
+      "learning_rate": 9.857996308936713e-07,
+      "loss": 0.4797,
+      "step": 11108
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2702463806136817,
+      "learning_rate": 9.847120655014032e-07,
+      "loss": 0.5111,
+      "step": 11109
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1980189183035805,
+      "learning_rate": 9.836250692907745e-07,
+      "loss": 0.471,
+      "step": 11110
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1265073850062017,
+      "learning_rate": 9.825386423304162e-07,
+      "loss": 0.4891,
+      "step": 11111
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.192428000653777,
+      "learning_rate": 9.814527846889165e-07,
+      "loss": 0.5272,
+      "step": 11112
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0941625634431469,
+      "learning_rate": 9.80367496434831e-07,
+      "loss": 0.4911,
+      "step": 11113
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.319738138985596,
+      "learning_rate": 9.792827776366797e-07,
+      "loss": 0.5725,
+      "step": 11114
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1911902560196133,
+      "learning_rate": 9.781986283629484e-07,
+      "loss": 0.4553,
+      "step": 11115
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2118604034914062,
+      "learning_rate": 9.771150486820857e-07,
+      "loss": 0.4901,
+      "step": 11116
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1574768053371496,
+      "learning_rate": 9.760320386625e-07,
+      "loss": 0.4629,
+      "step": 11117
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1492088996885166,
+      "learning_rate": 9.749495983725688e-07,
+      "loss": 0.3864,
+      "step": 11118
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.292519113779314,
+      "learning_rate": 9.73867727880633e-07,
+      "loss": 0.4772,
+      "step": 11119
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.204174277777636,
+      "learning_rate": 9.72786427254996e-07,
+      "loss": 0.5069,
+      "step": 11120
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1778397394371976,
+      "learning_rate": 9.717056965639281e-07,
+      "loss": 0.4645,
+      "step": 11121
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3041749050054863,
+      "learning_rate": 9.7062553587566e-07,
+      "loss": 0.5567,
+      "step": 11122
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.21181072621425,
+      "learning_rate": 9.695459452583843e-07,
+      "loss": 0.5382,
+      "step": 11123
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.22384610033517,
+      "learning_rate": 9.684669247802647e-07,
+      "loss": 0.5001,
+      "step": 11124
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.136770869892159,
+      "learning_rate": 9.673884745094253e-07,
+      "loss": 0.4699,
+      "step": 11125
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.175447841277305,
+      "learning_rate": 9.66310594513955e-07,
+      "loss": 0.4949,
+      "step": 11126
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.231113180313904,
+      "learning_rate": 9.652332848619027e-07,
+      "loss": 0.5281,
+      "step": 11127
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2594876241375021,
+      "learning_rate": 9.641565456212864e-07,
+      "loss": 0.4885,
+      "step": 11128
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2586171674296198,
+      "learning_rate": 9.63080376860086e-07,
+      "loss": 0.4656,
+      "step": 11129
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2159352198461497,
+      "learning_rate": 9.620047786462461e-07,
+      "loss": 0.4531,
+      "step": 11130
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1642756426974885,
+      "learning_rate": 9.609297510476767e-07,
+      "loss": 0.4431,
+      "step": 11131
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1896148721997954,
+      "learning_rate": 9.59855294132247e-07,
+      "loss": 0.4788,
+      "step": 11132
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0920543605876847,
+      "learning_rate": 9.587814079677915e-07,
+      "loss": 0.4649,
+      "step": 11133
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2793039103307267,
+      "learning_rate": 9.577080926221127e-07,
+      "loss": 0.5307,
+      "step": 11134
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2251348265584732,
+      "learning_rate": 9.566353481629742e-07,
+      "loss": 0.4814,
+      "step": 11135
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0523110637388777,
+      "learning_rate": 9.55563174658105e-07,
+      "loss": 0.4391,
+      "step": 11136
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2141699685702245,
+      "learning_rate": 9.544915721751946e-07,
+      "loss": 0.5207,
+      "step": 11137
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1176559959783576,
+      "learning_rate": 9.534205407818997e-07,
+      "loss": 0.5235,
+      "step": 11138
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3219456701123153,
+      "learning_rate": 9.523500805458408e-07,
+      "loss": 0.5502,
+      "step": 11139
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1973584685179508,
+      "learning_rate": 9.512801915346004e-07,
+      "loss": 0.5235,
+      "step": 11140
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2014360338980115,
+      "learning_rate": 9.502108738157279e-07,
+      "loss": 0.4954,
+      "step": 11141
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2806681265394901,
+      "learning_rate": 9.491421274567348e-07,
+      "loss": 0.5363,
+      "step": 11142
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.293565410769431,
+      "learning_rate": 9.480739525250938e-07,
+      "loss": 0.4993,
+      "step": 11143
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.213010920212061,
+      "learning_rate": 9.470063490882453e-07,
+      "loss": 0.4709,
+      "step": 11144
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.332143202744742,
+      "learning_rate": 9.459393172135934e-07,
+      "loss": 0.5443,
+      "step": 11145
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1991046764282272,
+      "learning_rate": 9.448728569685073e-07,
+      "loss": 0.4954,
+      "step": 11146
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1676084032724046,
+      "learning_rate": 9.438069684203144e-07,
+      "loss": 0.4936,
+      "step": 11147
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3055712967232531,
+      "learning_rate": 9.427416516363108e-07,
+      "loss": 0.5068,
+      "step": 11148
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1371821347136535,
+      "learning_rate": 9.416769066837561e-07,
+      "loss": 0.527,
+      "step": 11149
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.213476312975917,
+      "learning_rate": 9.406127336298731e-07,
+      "loss": 0.5763,
+      "step": 11150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1345408967247246,
+      "learning_rate": 9.395491325418505e-07,
+      "loss": 0.4662,
+      "step": 11151
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.124287485615347,
+      "learning_rate": 9.384861034868376e-07,
+      "loss": 0.4687,
+      "step": 11152
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0808653902659178,
+      "learning_rate": 9.374236465319453e-07,
+      "loss": 0.4809,
+      "step": 11153
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.186661725560806,
+      "learning_rate": 9.363617617442555e-07,
+      "loss": 0.453,
+      "step": 11154
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1402385004589048,
+      "learning_rate": 9.353004491908102e-07,
+      "loss": 0.4664,
+      "step": 11155
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.204701072527355,
+      "learning_rate": 9.342397089386168e-07,
+      "loss": 0.4924,
+      "step": 11156
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1756724056925454,
+      "learning_rate": 9.331795410546418e-07,
+      "loss": 0.4596,
+      "step": 11157
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2837324052475951,
+      "learning_rate": 9.321199456058205e-07,
+      "loss": 0.504,
+      "step": 11158
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1566139370133486,
+      "learning_rate": 9.310609226590516e-07,
+      "loss": 0.5008,
+      "step": 11159
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1717083621756874,
+      "learning_rate": 9.300024722811973e-07,
+      "loss": 0.5294,
+      "step": 11160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1600006718469187,
+      "learning_rate": 9.289445945390829e-07,
+      "loss": 0.4973,
+      "step": 11161
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.264032704458381,
+      "learning_rate": 9.278872894994962e-07,
+      "loss": 0.5283,
+      "step": 11162
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.167258515553021,
+      "learning_rate": 9.268305572291892e-07,
+      "loss": 0.4497,
+      "step": 11163
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2316437938684506,
+      "learning_rate": 9.257743977948808e-07,
+      "loss": 0.4425,
+      "step": 11164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1928585518713626,
+      "learning_rate": 9.247188112632522e-07,
+      "loss": 0.5146,
+      "step": 11165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2915489491790895,
+      "learning_rate": 9.236637977009466e-07,
+      "loss": 0.5629,
+      "step": 11166
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1797096806139902,
+      "learning_rate": 9.226093571745753e-07,
+      "loss": 0.4874,
+      "step": 11167
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2036613160736935,
+      "learning_rate": 9.215554897507062e-07,
+      "loss": 0.5092,
+      "step": 11168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.246713992170833,
+      "learning_rate": 9.205021954958781e-07,
+      "loss": 0.4835,
+      "step": 11169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2867357930373173,
+      "learning_rate": 9.194494744765902e-07,
+      "loss": 0.5126,
+      "step": 11170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.396936175481047,
+      "learning_rate": 9.183973267593083e-07,
+      "loss": 0.5147,
+      "step": 11171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2231216276236363,
+      "learning_rate": 9.173457524104579e-07,
+      "loss": 0.5107,
+      "step": 11172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1079604573303279,
+      "learning_rate": 9.162947514964283e-07,
+      "loss": 0.4855,
+      "step": 11173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2296632582566496,
+      "learning_rate": 9.152443240835774e-07,
+      "loss": 0.4597,
+      "step": 11174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1822916050489412,
+      "learning_rate": 9.141944702382233e-07,
+      "loss": 0.4939,
+      "step": 11175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.15621984287834,
+      "learning_rate": 9.131451900266497e-07,
+      "loss": 0.5091,
+      "step": 11176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2923908152108374,
+      "learning_rate": 9.120964835151025e-07,
+      "loss": 0.5664,
+      "step": 11177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1683436783909869,
+      "learning_rate": 9.110483507697909e-07,
+      "loss": 0.4956,
+      "step": 11178
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.143494380465078,
+      "learning_rate": 9.100007918568898e-07,
+      "loss": 0.459,
+      "step": 11179
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2018140611360189,
+      "learning_rate": 9.089538068425375e-07,
+      "loss": 0.5266,
+      "step": 11180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1874766598465667,
+      "learning_rate": 9.079073957928353e-07,
+      "loss": 0.4988,
+      "step": 11181
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.208609330237781,
+      "learning_rate": 9.068615587738495e-07,
+      "loss": 0.5083,
+      "step": 11182
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2322878999026405,
+      "learning_rate": 9.058162958516059e-07,
+      "loss": 0.4758,
+      "step": 11183
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2208875348988535,
+      "learning_rate": 9.047716070920987e-07,
+      "loss": 0.4875,
+      "step": 11184
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2170034756700947,
+      "learning_rate": 9.037274925612849e-07,
+      "loss": 0.5116,
+      "step": 11185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2934713164851488,
+      "learning_rate": 9.026839523250863e-07,
+      "loss": 0.5082,
+      "step": 11186
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3123901866205288,
+      "learning_rate": 9.016409864493869e-07,
+      "loss": 0.545,
+      "step": 11187
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.26818132669297,
+      "learning_rate": 9.005985950000318e-07,
+      "loss": 0.5059,
+      "step": 11188
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1834738390520687,
+      "learning_rate": 8.995567780428338e-07,
+      "loss": 0.4754,
+      "step": 11189
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1995655266695722,
+      "learning_rate": 8.985155356435704e-07,
+      "loss": 0.5217,
+      "step": 11190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2144283062695491,
+      "learning_rate": 8.974748678679768e-07,
+      "loss": 0.4439,
+      "step": 11191
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1772206769277709,
+      "learning_rate": 8.964347747817603e-07,
+      "loss": 0.529,
+      "step": 11192
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.199213729997,
+      "learning_rate": 8.953952564505819e-07,
+      "loss": 0.4795,
+      "step": 11193
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1775321723608596,
+      "learning_rate": 8.943563129400756e-07,
+      "loss": 0.4613,
+      "step": 11194
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1814972089007225,
+      "learning_rate": 8.933179443158336e-07,
+      "loss": 0.5584,
+      "step": 11195
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2351378003358497,
+      "learning_rate": 8.922801506434131e-07,
+      "loss": 0.5025,
+      "step": 11196
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3081507601983309,
+      "learning_rate": 8.912429319883398e-07,
+      "loss": 0.4902,
+      "step": 11197
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1408358862332597,
+      "learning_rate": 8.902062884160922e-07,
+      "loss": 0.4499,
+      "step": 11198
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1992552313707423,
+      "learning_rate": 8.891702199921226e-07,
+      "loss": 0.4653,
+      "step": 11199
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2486846679154369,
+      "learning_rate": 8.881347267818441e-07,
+      "loss": 0.4959,
+      "step": 11200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.136180055267908,
+      "learning_rate": 8.8709980885063e-07,
+      "loss": 0.4721,
+      "step": 11201
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2635046068476634,
+      "learning_rate": 8.860654662638235e-07,
+      "loss": 0.5093,
+      "step": 11202
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.206876627576729,
+      "learning_rate": 8.850316990867236e-07,
+      "loss": 0.4926,
+      "step": 11203
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2964802796246564,
+      "learning_rate": 8.839985073845991e-07,
+      "loss": 0.5135,
+      "step": 11204
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2375531387479795,
+      "learning_rate": 8.829658912226813e-07,
+      "loss": 0.5201,
+      "step": 11205
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1901414505052301,
+      "learning_rate": 8.819338506661646e-07,
+      "loss": 0.4904,
+      "step": 11206
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2286828538277166,
+      "learning_rate": 8.80902385780209e-07,
+      "loss": 0.4874,
+      "step": 11207
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3429025594877946,
+      "learning_rate": 8.798714966299327e-07,
+      "loss": 0.5427,
+      "step": 11208
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1468429394640147,
+      "learning_rate": 8.788411832804223e-07,
+      "loss": 0.5172,
+      "step": 11209
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.304672697976943,
+      "learning_rate": 8.77811445796728e-07,
+      "loss": 0.501,
+      "step": 11210
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1765707919979842,
+      "learning_rate": 8.767822842438601e-07,
+      "loss": 0.4648,
+      "step": 11211
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1916922304216908,
+      "learning_rate": 8.757536986867987e-07,
+      "loss": 0.5106,
+      "step": 11212
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1705712250162306,
+      "learning_rate": 8.747256891904787e-07,
+      "loss": 0.4985,
+      "step": 11213
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2584556214908422,
+      "learning_rate": 8.736982558198059e-07,
+      "loss": 0.4938,
+      "step": 11214
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1309118176358817,
+      "learning_rate": 8.726713986396484e-07,
+      "loss": 0.4599,
+      "step": 11215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2241311242172317,
+      "learning_rate": 8.716451177148355e-07,
+      "loss": 0.4867,
+      "step": 11216
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.108570451874717,
+      "learning_rate": 8.706194131101653e-07,
+      "loss": 0.4942,
+      "step": 11217
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2683000433801723,
+      "learning_rate": 8.695942848903905e-07,
+      "loss": 0.5085,
+      "step": 11218
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.168563844267776,
+      "learning_rate": 8.685697331202348e-07,
+      "loss": 0.4712,
+      "step": 11219
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1786564748803117,
+      "learning_rate": 8.675457578643865e-07,
+      "loss": 0.4629,
+      "step": 11220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2282432169230988,
+      "learning_rate": 8.665223591874894e-07,
+      "loss": 0.5262,
+      "step": 11221
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1806366493696432,
+      "learning_rate": 8.654995371541585e-07,
+      "loss": 0.5021,
+      "step": 11222
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2780642562645959,
+      "learning_rate": 8.644772918289723e-07,
+      "loss": 0.5358,
+      "step": 11223
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2631192303374286,
+      "learning_rate": 8.634556232764646e-07,
+      "loss": 0.5371,
+      "step": 11224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2665903624757802,
+      "learning_rate": 8.624345315611427e-07,
+      "loss": 0.5115,
+      "step": 11225
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2488369776436161,
+      "learning_rate": 8.614140167474716e-07,
+      "loss": 0.5275,
+      "step": 11226
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2711884935141544,
+      "learning_rate": 8.603940788998832e-07,
+      "loss": 0.5312,
+      "step": 11227
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2271019302193817,
+      "learning_rate": 8.593747180827728e-07,
+      "loss": 0.5376,
+      "step": 11228
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2257255079579459,
+      "learning_rate": 8.58355934360493e-07,
+      "loss": 0.536,
+      "step": 11229
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.148538598002825,
+      "learning_rate": 8.573377277973704e-07,
+      "loss": 0.451,
+      "step": 11230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2276454930989134,
+      "learning_rate": 8.563200984576847e-07,
+      "loss": 0.525,
+      "step": 11231
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1447443845019263,
+      "learning_rate": 8.553030464056867e-07,
+      "loss": 0.4689,
+      "step": 11232
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1342899553350756,
+      "learning_rate": 8.542865717055904e-07,
+      "loss": 0.485,
+      "step": 11233
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.17823768097095,
+      "learning_rate": 8.532706744215657e-07,
+      "loss": 0.5005,
+      "step": 11234
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.26999996042627,
+      "learning_rate": 8.522553546177536e-07,
+      "loss": 0.506,
+      "step": 11235
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.209582204096564,
+      "learning_rate": 8.512406123582583e-07,
+      "loss": 0.4593,
+      "step": 11236
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2588033623545312,
+      "learning_rate": 8.502264477071442e-07,
+      "loss": 0.538,
+      "step": 11237
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3303412402657289,
+      "learning_rate": 8.492128607284434e-07,
+      "loss": 0.5367,
+      "step": 11238
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.206029630701883,
+      "learning_rate": 8.481998514861434e-07,
+      "loss": 0.4401,
+      "step": 11239
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0809843321307926,
+      "learning_rate": 8.471874200442066e-07,
+      "loss": 0.4749,
+      "step": 11240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2413176363892566,
+      "learning_rate": 8.461755664665483e-07,
+      "loss": 0.524,
+      "step": 11241
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.144646752684332,
+      "learning_rate": 8.451642908170544e-07,
+      "loss": 0.4626,
+      "step": 11242
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.156760077474445,
+      "learning_rate": 8.441535931595735e-07,
+      "loss": 0.5196,
+      "step": 11243
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.30150829078651,
+      "learning_rate": 8.431434735579113e-07,
+      "loss": 0.5514,
+      "step": 11244
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2666890416344163,
+      "learning_rate": 8.421339320758459e-07,
+      "loss": 0.5174,
+      "step": 11245
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1652637914051374,
+      "learning_rate": 8.411249687771128e-07,
+      "loss": 0.4734,
+      "step": 11246
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2255565138244648,
+      "learning_rate": 8.401165837254144e-07,
+      "loss": 0.4757,
+      "step": 11247
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1574821608386052,
+      "learning_rate": 8.391087769844164e-07,
+      "loss": 0.4636,
+      "step": 11248
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.239983192299173,
+      "learning_rate": 8.381015486177446e-07,
+      "loss": 0.5103,
+      "step": 11249
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1932116727685784,
+      "learning_rate": 8.370948986889915e-07,
+      "loss": 0.4919,
+      "step": 11250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1838144027447353,
+      "learning_rate": 8.360888272617107e-07,
+      "loss": 0.5026,
+      "step": 11251
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6026362099686522,
+      "learning_rate": 8.350833343994225e-07,
+      "loss": 0.4943,
+      "step": 11252
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2231686039609668,
+      "learning_rate": 8.340784201656094e-07,
+      "loss": 0.5212,
+      "step": 11253
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.226911167227712,
+      "learning_rate": 8.330740846237128e-07,
+      "loss": 0.4926,
+      "step": 11254
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3333723638702852,
+      "learning_rate": 8.320703278371456e-07,
+      "loss": 0.5012,
+      "step": 11255
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1544069086032005,
+      "learning_rate": 8.31067149869279e-07,
+      "loss": 0.4843,
+      "step": 11256
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2797548828270313,
+      "learning_rate": 8.300645507834481e-07,
+      "loss": 0.5283,
+      "step": 11257
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1862572642225533,
+      "learning_rate": 8.290625306429545e-07,
+      "loss": 0.4402,
+      "step": 11258
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2463512573123212,
+      "learning_rate": 8.280610895110575e-07,
+      "loss": 0.5039,
+      "step": 11259
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2776910152326302,
+      "learning_rate": 8.270602274509864e-07,
+      "loss": 0.5541,
+      "step": 11260
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2480170734411393,
+      "learning_rate": 8.260599445259276e-07,
+      "loss": 0.5114,
+      "step": 11261
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.196292500393657,
+      "learning_rate": 8.250602407990361e-07,
+      "loss": 0.4747,
+      "step": 11262
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1423940657036784,
+      "learning_rate": 8.2406111633343e-07,
+      "loss": 0.4607,
+      "step": 11263
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1827345143338315,
+      "learning_rate": 8.230625711921858e-07,
+      "loss": 0.5077,
+      "step": 11264
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3033689128835662,
+      "learning_rate": 8.220646054383475e-07,
+      "loss": 0.5258,
+      "step": 11265
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.171313291079321,
+      "learning_rate": 8.210672191349222e-07,
+      "loss": 0.466,
+      "step": 11266
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1957569168787214,
+      "learning_rate": 8.20070412344881e-07,
+      "loss": 0.4996,
+      "step": 11267
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2988797514557535,
+      "learning_rate": 8.19074185131159e-07,
+      "loss": 0.5458,
+      "step": 11268
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1304129607189908,
+      "learning_rate": 8.180785375566491e-07,
+      "loss": 0.4761,
+      "step": 11269
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1488017847387322,
+      "learning_rate": 8.170834696842156e-07,
+      "loss": 0.4788,
+      "step": 11270
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1078762084548042,
+      "learning_rate": 8.160889815766782e-07,
+      "loss": 0.4699,
+      "step": 11271
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1308643821908522,
+      "learning_rate": 8.150950732968255e-07,
+      "loss": 0.4765,
+      "step": 11272
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.298289756769988,
+      "learning_rate": 8.141017449074096e-07,
+      "loss": 0.497,
+      "step": 11273
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2390328421777432,
+      "learning_rate": 8.131089964711447e-07,
+      "loss": 0.4963,
+      "step": 11274
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1554863444156502,
+      "learning_rate": 8.121168280507053e-07,
+      "loss": 0.507,
+      "step": 11275
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.170073994473048,
+      "learning_rate": 8.111252397087344e-07,
+      "loss": 0.5472,
+      "step": 11276
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2045351162134736,
+      "learning_rate": 8.101342315078342e-07,
+      "loss": 0.5229,
+      "step": 11277
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2169387269342302,
+      "learning_rate": 8.091438035105747e-07,
+      "loss": 0.4822,
+      "step": 11278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.243795733353835,
+      "learning_rate": 8.08153955779487e-07,
+      "loss": 0.5003,
+      "step": 11279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1594836921844616,
+      "learning_rate": 8.071646883770634e-07,
+      "loss": 0.4696,
+      "step": 11280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2186486617774832,
+      "learning_rate": 8.061760013657605e-07,
+      "loss": 0.5129,
+      "step": 11281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1750611289335768,
+      "learning_rate": 8.051878948080006e-07,
+      "loss": 0.5128,
+      "step": 11282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2175840033155843,
+      "learning_rate": 8.042003687661671e-07,
+      "loss": 0.5118,
+      "step": 11283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.33371138676787,
+      "learning_rate": 8.032134233026101e-07,
+      "loss": 0.5344,
+      "step": 11284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1675892088827573,
+      "learning_rate": 8.022270584796376e-07,
+      "loss": 0.4784,
+      "step": 11285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1941431846004729,
+      "learning_rate": 8.012412743595255e-07,
+      "loss": 0.5054,
+      "step": 11286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2484860311726615,
+      "learning_rate": 8.002560710045115e-07,
+      "loss": 0.5308,
+      "step": 11287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2217283292580965,
+      "learning_rate": 7.992714484767949e-07,
+      "loss": 0.5337,
+      "step": 11288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.147829634499345,
+      "learning_rate": 7.982874068385438e-07,
+      "loss": 0.4634,
+      "step": 11289
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.218583706979347,
+      "learning_rate": 7.973039461518827e-07,
+      "loss": 0.4937,
+      "step": 11290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.302723302608259,
+      "learning_rate": 7.963210664789022e-07,
+      "loss": 0.5007,
+      "step": 11291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1091457721128348,
+      "learning_rate": 7.953387678816571e-07,
+      "loss": 0.4481,
+      "step": 11292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.177266700213032,
+      "learning_rate": 7.943570504221654e-07,
+      "loss": 0.4766,
+      "step": 11293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0073171773452696,
+      "learning_rate": 7.933759141624098e-07,
+      "loss": 0.4559,
+      "step": 11294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1714233036537556,
+      "learning_rate": 7.923953591643308e-07,
+      "loss": 0.4675,
+      "step": 11295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1046513207539383,
+      "learning_rate": 7.914153854898376e-07,
+      "loss": 0.4606,
+      "step": 11296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1417194498116632,
+      "learning_rate": 7.90435993200801e-07,
+      "loss": 0.4495,
+      "step": 11297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.271493422421848,
+      "learning_rate": 7.89457182359058e-07,
+      "loss": 0.516,
+      "step": 11298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.261224750383257,
+      "learning_rate": 7.884789530264004e-07,
+      "loss": 0.507,
+      "step": 11299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2302893613810355,
+      "learning_rate": 7.875013052645941e-07,
+      "loss": 0.5111,
+      "step": 11300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0920654403305488,
+      "learning_rate": 7.865242391353589e-07,
+      "loss": 0.4337,
+      "step": 11301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0906805838600118,
+      "learning_rate": 7.855477547003831e-07,
+      "loss": 0.4428,
+      "step": 11302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2393748751903324,
+      "learning_rate": 7.845718520213186e-07,
+      "loss": 0.5743,
+      "step": 11303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1909548044861884,
+      "learning_rate": 7.835965311597804e-07,
+      "loss": 0.4726,
+      "step": 11304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.138015269855835,
+      "learning_rate": 7.826217921773416e-07,
+      "loss": 0.4813,
+      "step": 11305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2169900070413333,
+      "learning_rate": 7.81647635135544e-07,
+      "loss": 0.501,
+      "step": 11306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3549569782592175,
+      "learning_rate": 7.806740600958918e-07,
+      "loss": 0.5692,
+      "step": 11307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1757004415428356,
+      "learning_rate": 7.797010671198534e-07,
+      "loss": 0.4679,
+      "step": 11308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.242894433351586,
+      "learning_rate": 7.787286562688556e-07,
+      "loss": 0.4583,
+      "step": 11309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2404504782967962,
+      "learning_rate": 7.777568276042946e-07,
+      "loss": 0.5195,
+      "step": 11310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1420621831903188,
+      "learning_rate": 7.767855811875236e-07,
+      "loss": 0.4584,
+      "step": 11311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3082915455326396,
+      "learning_rate": 7.758149170798656e-07,
+      "loss": 0.5237,
+      "step": 11312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1417436731546178,
+      "learning_rate": 7.748448353426019e-07,
+      "loss": 0.4933,
+      "step": 11313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2384882131238193,
+      "learning_rate": 7.7387533603698e-07,
+      "loss": 0.5721,
+      "step": 11314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2964811531341784,
+      "learning_rate": 7.729064192242075e-07,
+      "loss": 0.5366,
+      "step": 11315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1619825277902553,
+      "learning_rate": 7.71938084965459e-07,
+      "loss": 0.4861,
+      "step": 11316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1922181911692673,
+      "learning_rate": 7.709703333218698e-07,
+      "loss": 0.498,
+      "step": 11317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2431795489357234,
+      "learning_rate": 7.700031643545402e-07,
+      "loss": 0.5175,
+      "step": 11318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2842321348902581,
+      "learning_rate": 7.690365781245291e-07,
+      "loss": 0.532,
+      "step": 11319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1670132814679615,
+      "learning_rate": 7.680705746928663e-07,
+      "loss": 0.485,
+      "step": 11320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1312991326553257,
+      "learning_rate": 7.671051541205376e-07,
+      "loss": 0.482,
+      "step": 11321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.234337818213493,
+      "learning_rate": 7.661403164684955e-07,
+      "loss": 0.5314,
+      "step": 11322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0515430057526833,
+      "learning_rate": 7.651760617976556e-07,
+      "loss": 0.4569,
+      "step": 11323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.062439187497396,
+      "learning_rate": 7.64212390168898e-07,
+      "loss": 0.4803,
+      "step": 11324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1316083624984197,
+      "learning_rate": 7.63249301643062e-07,
+      "loss": 0.4445,
+      "step": 11325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0981880610189232,
+      "learning_rate": 7.622867962809521e-07,
+      "loss": 0.4418,
+      "step": 11326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4834911726794078,
+      "learning_rate": 7.613248741433365e-07,
+      "loss": 0.5421,
+      "step": 11327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.297139519980659,
+      "learning_rate": 7.603635352909489e-07,
+      "loss": 0.5212,
+      "step": 11328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2572005779620647,
+      "learning_rate": 7.594027797844805e-07,
+      "loss": 0.5428,
+      "step": 11329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1773678033570112,
+      "learning_rate": 7.584426076845908e-07,
+      "loss": 0.5424,
+      "step": 11330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.211774475272321,
+      "learning_rate": 7.574830190518978e-07,
+      "loss": 0.5016,
+      "step": 11331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1955396149731525,
+      "learning_rate": 7.565240139469877e-07,
+      "loss": 0.5345,
+      "step": 11332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1980447894147563,
+      "learning_rate": 7.555655924304062e-07,
+      "loss": 0.5159,
+      "step": 11333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.16401743961654,
+      "learning_rate": 7.54607754562664e-07,
+      "loss": 0.5148,
+      "step": 11334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2553314000920262,
+      "learning_rate": 7.536505004042361e-07,
+      "loss": 0.5302,
+      "step": 11335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2720684813502816,
+      "learning_rate": 7.526938300155539e-07,
+      "loss": 0.5237,
+      "step": 11336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1479334860583175,
+      "learning_rate": 7.517377434570217e-07,
+      "loss": 0.4627,
+      "step": 11337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2516650078140692,
+      "learning_rate": 7.50782240789002e-07,
+      "loss": 0.4981,
+      "step": 11338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2217910191390557,
+      "learning_rate": 7.498273220718167e-07,
+      "loss": 0.5245,
+      "step": 11339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.290836708494584,
+      "learning_rate": 7.488729873657586e-07,
+      "loss": 0.5503,
+      "step": 11340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.343059761731569,
+      "learning_rate": 7.479192367310773e-07,
+      "loss": 0.4808,
+      "step": 11341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.221843266379924,
+      "learning_rate": 7.46966070227989e-07,
+      "loss": 0.4829,
+      "step": 11342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2567819199887884,
+      "learning_rate": 7.460134879166725e-07,
+      "loss": 0.5545,
+      "step": 11343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1358596334306892,
+      "learning_rate": 7.450614898572683e-07,
+      "loss": 0.5319,
+      "step": 11344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2651083562830525,
+      "learning_rate": 7.44110076109883e-07,
+      "loss": 0.4735,
+      "step": 11345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1864908849701237,
+      "learning_rate": 7.431592467345816e-07,
+      "loss": 0.4745,
+      "step": 11346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2582502373464548,
+      "learning_rate": 7.422090017913952e-07,
+      "loss": 0.5295,
+      "step": 11347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2223549858629155,
+      "learning_rate": 7.412593413403202e-07,
+      "loss": 0.4948,
+      "step": 11348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2842134304472528,
+      "learning_rate": 7.403102654413108e-07,
+      "loss": 0.5468,
+      "step": 11349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1531296975471728,
+      "learning_rate": 7.393617741542891e-07,
+      "loss": 0.4962,
+      "step": 11350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.197141158586891,
+      "learning_rate": 7.384138675391362e-07,
+      "loss": 0.434,
+      "step": 11351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2516980082367521,
+      "learning_rate": 7.374665456556984e-07,
+      "loss": 0.519,
+      "step": 11352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.252221089220329,
+      "learning_rate": 7.365198085637871e-07,
+      "loss": 0.5258,
+      "step": 11353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3108531065409659,
+      "learning_rate": 7.35573656323173e-07,
+      "loss": 0.4826,
+      "step": 11354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2326545797910435,
+      "learning_rate": 7.346280889935931e-07,
+      "loss": 0.4995,
+      "step": 11355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1569810051475684,
+      "learning_rate": 7.336831066347438e-07,
+      "loss": 0.4898,
+      "step": 11356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1640434519598215,
+      "learning_rate": 7.327387093062887e-07,
+      "loss": 0.4469,
+      "step": 11357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1720978080017563,
+      "learning_rate": 7.317948970678524e-07,
+      "loss": 0.483,
+      "step": 11358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1839182191777131,
+      "learning_rate": 7.308516699790202e-07,
+      "loss": 0.4855,
+      "step": 11359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.253123529295403,
+      "learning_rate": 7.29909028099347e-07,
+      "loss": 0.4706,
+      "step": 11360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.188400027867705,
+      "learning_rate": 7.289669714883419e-07,
+      "loss": 0.5155,
+      "step": 11361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.222107639263959,
+      "learning_rate": 7.280255002054848e-07,
+      "loss": 0.5247,
+      "step": 11362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.186555034703382,
+      "learning_rate": 7.270846143102139e-07,
+      "loss": 0.4576,
+      "step": 11363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2188331869052162,
+      "learning_rate": 7.26144313861934e-07,
+      "loss": 0.5334,
+      "step": 11364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3326202611707463,
+      "learning_rate": 7.252045989200118e-07,
+      "loss": 0.5555,
+      "step": 11365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3235553441280918,
+      "learning_rate": 7.242654695437734e-07,
+      "loss": 0.5047,
+      "step": 11366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1612001596848196,
+      "learning_rate": 7.233269257925124e-07,
+      "loss": 0.4436,
+      "step": 11367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2319523178292437,
+      "learning_rate": 7.223889677254858e-07,
+      "loss": 0.4354,
+      "step": 11368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.183256397074468,
+      "learning_rate": 7.214515954019086e-07,
+      "loss": 0.5105,
+      "step": 11369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.231004725390614,
+      "learning_rate": 7.205148088809632e-07,
+      "loss": 0.476,
+      "step": 11370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2966763907331895,
+      "learning_rate": 7.195786082217937e-07,
+      "loss": 0.5428,
+      "step": 11371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.314128138369065,
+      "learning_rate": 7.18642993483507e-07,
+      "loss": 0.5137,
+      "step": 11372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1548414156155586,
+      "learning_rate": 7.177079647251728e-07,
+      "loss": 0.4577,
+      "step": 11373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.254703779103097,
+      "learning_rate": 7.167735220058258e-07,
+      "loss": 0.4953,
+      "step": 11374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0821017025482853,
+      "learning_rate": 7.158396653844635e-07,
+      "loss": 0.4574,
+      "step": 11375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2697560742182812,
+      "learning_rate": 7.149063949200408e-07,
+      "loss": 0.4829,
+      "step": 11376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.234020713471559,
+      "learning_rate": 7.139737106714817e-07,
+      "loss": 0.4451,
+      "step": 11377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2116252299853332,
+      "learning_rate": 7.130416126976747e-07,
+      "loss": 0.4725,
+      "step": 11378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2450852573264932,
+      "learning_rate": 7.121101010574626e-07,
+      "loss": 0.5025,
+      "step": 11379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2100191140832603,
+      "learning_rate": 7.111791758096609e-07,
+      "loss": 0.4665,
+      "step": 11380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1004965203423887,
+      "learning_rate": 7.102488370130411e-07,
+      "loss": 0.5057,
+      "step": 11381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.15057159192129,
+      "learning_rate": 7.093190847263398e-07,
+      "loss": 0.487,
+      "step": 11382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2147374244599554,
+      "learning_rate": 7.083899190082588e-07,
+      "loss": 0.4996,
+      "step": 11383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2225960421134476,
+      "learning_rate": 7.074613399174601e-07,
+      "loss": 0.4965,
+      "step": 11384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.282002809384581,
+      "learning_rate": 7.065333475125713e-07,
+      "loss": 0.4787,
+      "step": 11385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2490485862126859,
+      "learning_rate": 7.05605941852181e-07,
+      "loss": 0.544,
+      "step": 11386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2084481524581705,
+      "learning_rate": 7.046791229948391e-07,
+      "loss": 0.46,
+      "step": 11387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.161967754550203,
+      "learning_rate": 7.037528909990632e-07,
+      "loss": 0.4995,
+      "step": 11388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2209253215743834,
+      "learning_rate": 7.028272459233277e-07,
+      "loss": 0.5006,
+      "step": 11389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.216686017174321,
+      "learning_rate": 7.019021878260757e-07,
+      "loss": 0.5209,
+      "step": 11390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0741391239132188,
+      "learning_rate": 7.009777167657117e-07,
+      "loss": 0.4743,
+      "step": 11391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2367567913207473,
+      "learning_rate": 7.000538328006001e-07,
+      "loss": 0.502,
+      "step": 11392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2878968488307398,
+      "learning_rate": 6.99130535989071e-07,
+      "loss": 0.4929,
+      "step": 11393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2053002349094968,
+      "learning_rate": 6.982078263894176e-07,
+      "loss": 0.52,
+      "step": 11394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0935505821449385,
+      "learning_rate": 6.972857040598945e-07,
+      "loss": 0.4696,
+      "step": 11395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2263248966687565,
+      "learning_rate": 6.96364169058722e-07,
+      "loss": 0.5011,
+      "step": 11396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3132745183242291,
+      "learning_rate": 6.954432214440798e-07,
+      "loss": 0.5406,
+      "step": 11397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1288686039861384,
+      "learning_rate": 6.945228612741129e-07,
+      "loss": 0.4861,
+      "step": 11398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2181872022081492,
+      "learning_rate": 6.936030886069256e-07,
+      "loss": 0.5083,
+      "step": 11399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1324142282102119,
+      "learning_rate": 6.926839035005905e-07,
+      "loss": 0.4333,
+      "step": 11400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9944093112552583,
+      "learning_rate": 6.917653060131413e-07,
+      "loss": 0.4189,
+      "step": 11401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0954491087959555,
+      "learning_rate": 6.908472962025714e-07,
+      "loss": 0.4524,
+      "step": 11402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.415816341715057,
+      "learning_rate": 6.899298741268412e-07,
+      "loss": 0.531,
+      "step": 11403
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1697057355579827,
+      "learning_rate": 6.89013039843871e-07,
+      "loss": 0.4878,
+      "step": 11404
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2172217446285696,
+      "learning_rate": 6.880967934115457e-07,
+      "loss": 0.5402,
+      "step": 11405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1613533702458352,
+      "learning_rate": 6.871811348877144e-07,
+      "loss": 0.5041,
+      "step": 11406
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2884818416554964,
+      "learning_rate": 6.862660643301855e-07,
+      "loss": 0.5085,
+      "step": 11407
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0959561033820955,
+      "learning_rate": 6.853515817967327e-07,
+      "loss": 0.4653,
+      "step": 11408
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.234250074414797,
+      "learning_rate": 6.844376873450908e-07,
+      "loss": 0.5013,
+      "step": 11409
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1417181446600595,
+      "learning_rate": 6.835243810329595e-07,
+      "loss": 0.5016,
+      "step": 11410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1355340296199523,
+      "learning_rate": 6.826116629180024e-07,
+      "loss": 0.48,
+      "step": 11411
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1304767073197668,
+      "learning_rate": 6.816995330578413e-07,
+      "loss": 0.4767,
+      "step": 11412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3036671834739988,
+      "learning_rate": 6.807879915100646e-07,
+      "loss": 0.5729,
+      "step": 11413
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2731069124010923,
+      "learning_rate": 6.798770383322218e-07,
+      "loss": 0.5003,
+      "step": 11414
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1115256403832667,
+      "learning_rate": 6.789666735818279e-07,
+      "loss": 0.452,
+      "step": 11415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2920868508050043,
+      "learning_rate": 6.780568973163604e-07,
+      "loss": 0.5029,
+      "step": 11416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0904333782458138,
+      "learning_rate": 6.771477095932533e-07,
+      "loss": 0.4334,
+      "step": 11417
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1416562787625226,
+      "learning_rate": 6.762391104699129e-07,
+      "loss": 0.4941,
+      "step": 11418
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.265655517210204,
+      "learning_rate": 6.753311000036999e-07,
+      "loss": 0.54,
+      "step": 11419
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1419123348152156,
+      "learning_rate": 6.744236782519431e-07,
+      "loss": 0.5165,
+      "step": 11420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1049265795692982,
+      "learning_rate": 6.735168452719354e-07,
+      "loss": 0.5151,
+      "step": 11421
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2633155662391082,
+      "learning_rate": 6.726106011209266e-07,
+      "loss": 0.5072,
+      "step": 11422
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2010426521572934,
+      "learning_rate": 6.71704945856132e-07,
+      "loss": 0.5102,
+      "step": 11423
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3141656025503459,
+      "learning_rate": 6.707998795347326e-07,
+      "loss": 0.5289,
+      "step": 11424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2124094722813237,
+      "learning_rate": 6.698954022138692e-07,
+      "loss": 0.5087,
+      "step": 11425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2292741577648272,
+      "learning_rate": 6.689915139506475e-07,
+      "loss": 0.5242,
+      "step": 11426
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2817797612335684,
+      "learning_rate": 6.680882148021318e-07,
+      "loss": 0.5443,
+      "step": 11427
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2314015561466816,
+      "learning_rate": 6.67185504825355e-07,
+      "loss": 0.4935,
+      "step": 11428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2000672758635287,
+      "learning_rate": 6.662833840773064e-07,
+      "loss": 0.4746,
+      "step": 11429
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2415612999588386,
+      "learning_rate": 6.653818526149436e-07,
+      "loss": 0.5062,
+      "step": 11430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2564989422157289,
+      "learning_rate": 6.644809104951866e-07,
+      "loss": 0.496,
+      "step": 11431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.169186215924927,
+      "learning_rate": 6.635805577749133e-07,
+      "loss": 0.4704,
+      "step": 11432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2325190341090977,
+      "learning_rate": 6.626807945109681e-07,
+      "loss": 0.5077,
+      "step": 11433
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1688490902106825,
+      "learning_rate": 6.61781620760159e-07,
+      "loss": 0.5123,
+      "step": 11434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.349198230439379,
+      "learning_rate": 6.608830365792551e-07,
+      "loss": 0.5181,
+      "step": 11435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2908356002892785,
+      "learning_rate": 6.59985042024991e-07,
+      "loss": 0.5258,
+      "step": 11436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3881548685006135,
+      "learning_rate": 6.590876371540567e-07,
+      "loss": 0.5603,
+      "step": 11437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.184198961285118,
+      "learning_rate": 6.581908220231137e-07,
+      "loss": 0.4849,
+      "step": 11438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1663623083566974,
+      "learning_rate": 6.572945966887812e-07,
+      "loss": 0.4801,
+      "step": 11439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2436056139538596,
+      "learning_rate": 6.563989612076416e-07,
+      "loss": 0.5004,
+      "step": 11440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2734395126607705,
+      "learning_rate": 6.55503915636243e-07,
+      "loss": 0.4826,
+      "step": 11441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2667629636165563,
+      "learning_rate": 6.546094600310949e-07,
+      "loss": 0.5281,
+      "step": 11442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2414451636234511,
+      "learning_rate": 6.537155944486662e-07,
+      "loss": 0.5047,
+      "step": 11443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.306529634833789,
+      "learning_rate": 6.52822318945392e-07,
+      "loss": 0.5592,
+      "step": 11444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.172634488996047,
+      "learning_rate": 6.519296335776703e-07,
+      "loss": 0.5198,
+      "step": 11445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2214905175376964,
+      "learning_rate": 6.510375384018619e-07,
+      "loss": 0.4983,
+      "step": 11446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1749153370487948,
+      "learning_rate": 6.50146033474287e-07,
+      "loss": 0.473,
+      "step": 11447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2645459225613036,
+      "learning_rate": 6.492551188512331e-07,
+      "loss": 0.5368,
+      "step": 11448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2899787651946117,
+      "learning_rate": 6.483647945889449e-07,
+      "loss": 0.528,
+      "step": 11449
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3149651129703641,
+      "learning_rate": 6.474750607436364e-07,
+      "loss": 0.5215,
+      "step": 11450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1656019540916982,
+      "learning_rate": 6.465859173714784e-07,
+      "loss": 0.484,
+      "step": 11451
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3168021492525686,
+      "learning_rate": 6.456973645286113e-07,
+      "loss": 0.4718,
+      "step": 11452
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0959935202963917,
+      "learning_rate": 6.448094022711304e-07,
+      "loss": 0.4295,
+      "step": 11453
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2052575569794224,
+      "learning_rate": 6.439220306550975e-07,
+      "loss": 0.4882,
+      "step": 11454
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2282212819027056,
+      "learning_rate": 6.430352497365377e-07,
+      "loss": 0.5026,
+      "step": 11455
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2250761027927906,
+      "learning_rate": 6.421490595714408e-07,
+      "loss": 0.4626,
+      "step": 11456
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0775728332084265,
+      "learning_rate": 6.412634602157519e-07,
+      "loss": 0.5068,
+      "step": 11457
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0999708106763129,
+      "learning_rate": 6.403784517253863e-07,
+      "loss": 0.4958,
+      "step": 11458
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1620736252871733,
+      "learning_rate": 6.394940341562173e-07,
+      "loss": 0.4906,
+      "step": 11459
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1828295062718412,
+      "learning_rate": 6.386102075640843e-07,
+      "loss": 0.4696,
+      "step": 11460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2605761382468663,
+      "learning_rate": 6.377269720047863e-07,
+      "loss": 0.5236,
+      "step": 11461
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2409199419318881,
+      "learning_rate": 6.368443275340897e-07,
+      "loss": 0.5308,
+      "step": 11462
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2605009077099896,
+      "learning_rate": 6.359622742077165e-07,
+      "loss": 0.5165,
+      "step": 11463
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1387573524179118,
+      "learning_rate": 6.350808120813567e-07,
+      "loss": 0.4512,
+      "step": 11464
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3529922772657272,
+      "learning_rate": 6.341999412106625e-07,
+      "loss": 0.5448,
+      "step": 11465
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1773897239249582,
+      "learning_rate": 6.333196616512493e-07,
+      "loss": 0.5101,
+      "step": 11466
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2891223315600924,
+      "learning_rate": 6.324399734586884e-07,
+      "loss": 0.5234,
+      "step": 11467
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2360339063234378,
+      "learning_rate": 6.315608766885251e-07,
+      "loss": 0.4961,
+      "step": 11468
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1634495292572455,
+      "learning_rate": 6.306823713962563e-07,
+      "loss": 0.4774,
+      "step": 11469
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2369934509523381,
+      "learning_rate": 6.298044576373485e-07,
+      "loss": 0.5346,
+      "step": 11470
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1100206242612793,
+      "learning_rate": 6.289271354672299e-07,
+      "loss": 0.4141,
+      "step": 11471
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1974501599828717,
+      "learning_rate": 6.280504049412905e-07,
+      "loss": 0.5232,
+      "step": 11472
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2707960665298836,
+      "learning_rate": 6.271742661148806e-07,
+      "loss": 0.4956,
+      "step": 11473
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1678222258144697,
+      "learning_rate": 6.262987190433168e-07,
+      "loss": 0.4848,
+      "step": 11474
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3655513455498731,
+      "learning_rate": 6.254237637818772e-07,
+      "loss": 0.4862,
+      "step": 11475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2370121466272352,
+      "learning_rate": 6.245494003858022e-07,
+      "loss": 0.5202,
+      "step": 11476
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1478723186273423,
+      "learning_rate": 6.236756289102941e-07,
+      "loss": 0.4699,
+      "step": 11477
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.331742151831977,
+      "learning_rate": 6.2280244941052e-07,
+      "loss": 0.4812,
+      "step": 11478
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2106345997647863,
+      "learning_rate": 6.219298619416059e-07,
+      "loss": 0.5125,
+      "step": 11479
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.185470000387417,
+      "learning_rate": 6.210578665586442e-07,
+      "loss": 0.5052,
+      "step": 11480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.212580691926514,
+      "learning_rate": 6.201864633166877e-07,
+      "loss": 0.5194,
+      "step": 11481
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2551719005990398,
+      "learning_rate": 6.193156522707555e-07,
+      "loss": 0.486,
+      "step": 11482
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2418040037475517,
+      "learning_rate": 6.184454334758227e-07,
+      "loss": 0.5533,
+      "step": 11483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2998216084812626,
+      "learning_rate": 6.175758069868321e-07,
+      "loss": 0.5332,
+      "step": 11484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.230697416826139,
+      "learning_rate": 6.167067728586873e-07,
+      "loss": 0.5078,
+      "step": 11485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2159069842350292,
+      "learning_rate": 6.158383311462568e-07,
+      "loss": 0.5583,
+      "step": 11486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2315289970357401,
+      "learning_rate": 6.149704819043667e-07,
+      "loss": 0.5096,
+      "step": 11487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2246629134404277,
+      "learning_rate": 6.141032251878132e-07,
+      "loss": 0.4753,
+      "step": 11488
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.302875059386743,
+      "learning_rate": 6.132365610513457e-07,
+      "loss": 0.5786,
+      "step": 11489
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2691977194838018,
+      "learning_rate": 6.123704895496829e-07,
+      "loss": 0.4784,
+      "step": 11490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1189072184138322,
+      "learning_rate": 6.115050107375053e-07,
+      "loss": 0.4325,
+      "step": 11491
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2353978325543034,
+      "learning_rate": 6.106401246694549e-07,
+      "loss": 0.4636,
+      "step": 11492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2985534836510755,
+      "learning_rate": 6.097758314001379e-07,
+      "loss": 0.5072,
+      "step": 11493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2031456090970258,
+      "learning_rate": 6.089121309841173e-07,
+      "loss": 0.4634,
+      "step": 11494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1414363535958993,
+      "learning_rate": 6.080490234759262e-07,
+      "loss": 0.458,
+      "step": 11495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2074600495073577,
+      "learning_rate": 6.071865089300577e-07,
+      "loss": 0.4909,
+      "step": 11496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.059273983350523,
+      "learning_rate": 6.063245874009638e-07,
+      "loss": 0.4521,
+      "step": 11497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1210650397156736,
+      "learning_rate": 6.054632589430654e-07,
+      "loss": 0.484,
+      "step": 11498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2262033313266074,
+      "learning_rate": 6.0460252361074e-07,
+      "loss": 0.5371,
+      "step": 11499
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.251673198476342,
+      "learning_rate": 6.037423814583299e-07,
+      "loss": 0.5448,
+      "step": 11500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2883652622087698,
+      "learning_rate": 6.028828325401426e-07,
+      "loss": 0.5224,
+      "step": 11501
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2830469509641087,
+      "learning_rate": 6.020238769104447e-07,
+      "loss": 0.4772,
+      "step": 11502
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.263750312646663,
+      "learning_rate": 6.011655146234674e-07,
+      "loss": 0.5337,
+      "step": 11503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1592515699792645,
+      "learning_rate": 6.003077457334017e-07,
+      "loss": 0.4858,
+      "step": 11504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2008153370751402,
+      "learning_rate": 5.994505702944042e-07,
+      "loss": 0.5066,
+      "step": 11505
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2612124628959334,
+      "learning_rate": 5.98593988360594e-07,
+      "loss": 0.5015,
+      "step": 11506
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.147990340806951,
+      "learning_rate": 5.977379999860488e-07,
+      "loss": 0.4738,
+      "step": 11507
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.114273164795503,
+      "learning_rate": 5.968826052248145e-07,
+      "loss": 0.4502,
+      "step": 11508
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2976836934825,
+      "learning_rate": 5.960278041308931e-07,
+      "loss": 0.4902,
+      "step": 11509
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2014418880015592,
+      "learning_rate": 5.951735967582551e-07,
+      "loss": 0.5329,
+      "step": 11510
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2742328617289833,
+      "learning_rate": 5.943199831608304e-07,
+      "loss": 0.5202,
+      "step": 11511
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2587320982219765,
+      "learning_rate": 5.934669633925116e-07,
+      "loss": 0.4809,
+      "step": 11512
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.16595699092592,
+      "learning_rate": 5.926145375071568e-07,
+      "loss": 0.4664,
+      "step": 11513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1908485983601393,
+      "learning_rate": 5.917627055585807e-07,
+      "loss": 0.4664,
+      "step": 11514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.157351613430679,
+      "learning_rate": 5.909114676005645e-07,
+      "loss": 0.462,
+      "step": 11515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.277930635256316,
+      "learning_rate": 5.900608236868532e-07,
+      "loss": 0.5313,
+      "step": 11516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4648454589833781,
+      "learning_rate": 5.892107738711505e-07,
+      "loss": 0.5294,
+      "step": 11517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2096183236361553,
+      "learning_rate": 5.883613182071257e-07,
+      "loss": 0.5709,
+      "step": 11518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1865059054613942,
+      "learning_rate": 5.875124567484069e-07,
+      "loss": 0.5131,
+      "step": 11519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.148821137376876,
+      "learning_rate": 5.866641895485892e-07,
+      "loss": 0.4299,
+      "step": 11520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1251920430196,
+      "learning_rate": 5.858165166612273e-07,
+      "loss": 0.4955,
+      "step": 11521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.195691366552233,
+      "learning_rate": 5.849694381398396e-07,
+      "loss": 0.544,
+      "step": 11522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1546910060060207,
+      "learning_rate": 5.841229540379079e-07,
+      "loss": 0.5319,
+      "step": 11523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1479977135452735,
+      "learning_rate": 5.832770644088726e-07,
+      "loss": 0.5128,
+      "step": 11524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2128555857930658,
+      "learning_rate": 5.824317693061387e-07,
+      "loss": 0.4753,
+      "step": 11525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.241292619155997,
+      "learning_rate": 5.815870687830782e-07,
+      "loss": 0.5141,
+      "step": 11526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1653126397801563,
+      "learning_rate": 5.807429628930172e-07,
+      "loss": 0.4541,
+      "step": 11527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.177684523506373,
+      "learning_rate": 5.798994516892509e-07,
+      "loss": 0.5136,
+      "step": 11528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0739360090439753,
+      "learning_rate": 5.790565352250322e-07,
+      "loss": 0.4287,
+      "step": 11529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.301586646397559,
+      "learning_rate": 5.782142135535806e-07,
+      "loss": 0.4968,
+      "step": 11530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1061639655327724,
+      "learning_rate": 5.77372486728075e-07,
+      "loss": 0.4648,
+      "step": 11531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1582700250097908,
+      "learning_rate": 5.765313548016593e-07,
+      "loss": 0.5057,
+      "step": 11532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3245410808213256,
+      "learning_rate": 5.756908178274389e-07,
+      "loss": 0.5502,
+      "step": 11533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1218823042182844,
+      "learning_rate": 5.748508758584792e-07,
+      "loss": 0.4495,
+      "step": 11534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2571300289967215,
+      "learning_rate": 5.740115289478109e-07,
+      "loss": 0.4927,
+      "step": 11535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2030677781802523,
+      "learning_rate": 5.731727771484275e-07,
+      "loss": 0.5042,
+      "step": 11536
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1734218814790778,
+      "learning_rate": 5.723346205132818e-07,
+      "loss": 0.5223,
+      "step": 11537
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3448370595175896,
+      "learning_rate": 5.714970590952939e-07,
+      "loss": 0.5226,
+      "step": 11538
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0855007253791147,
+      "learning_rate": 5.706600929473382e-07,
+      "loss": 0.4566,
+      "step": 11539
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2138332076775586,
+      "learning_rate": 5.698237221222614e-07,
+      "loss": 0.5171,
+      "step": 11540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2539276881776915,
+      "learning_rate": 5.689879466728654e-07,
+      "loss": 0.5312,
+      "step": 11541
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3131047626662606,
+      "learning_rate": 5.681527666519182e-07,
+      "loss": 0.5372,
+      "step": 11542
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2661229319706675,
+      "learning_rate": 5.673181821121509e-07,
+      "loss": 0.4981,
+      "step": 11543
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1311052809400346,
+      "learning_rate": 5.664841931062504e-07,
+      "loss": 0.5265,
+      "step": 11544
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2112644861573485,
+      "learning_rate": 5.656507996868743e-07,
+      "loss": 0.4923,
+      "step": 11545
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1716073811446706,
+      "learning_rate": 5.648180019066385e-07,
+      "loss": 0.5145,
+      "step": 11546
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.033528786806913,
+      "learning_rate": 5.639857998181208e-07,
+      "loss": 0.4607,
+      "step": 11547
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.071128589778587,
+      "learning_rate": 5.631541934738627e-07,
+      "loss": 0.4024,
+      "step": 11548
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1293005865056507,
+      "learning_rate": 5.623231829263698e-07,
+      "loss": 0.4428,
+      "step": 11549
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0832866450788494,
+      "learning_rate": 5.614927682281046e-07,
+      "loss": 0.4554,
+      "step": 11550
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.252620239577666,
+      "learning_rate": 5.606629494314963e-07,
+      "loss": 0.4963,
+      "step": 11551
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3517702858025151,
+      "learning_rate": 5.598337265889375e-07,
+      "loss": 0.5377,
+      "step": 11552
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2697677626744661,
+      "learning_rate": 5.590050997527808e-07,
+      "loss": 0.4895,
+      "step": 11553
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0985110067025359,
+      "learning_rate": 5.581770689753429e-07,
+      "loss": 0.438,
+      "step": 11554
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1086773895262008,
+      "learning_rate": 5.57349634308898e-07,
+      "loss": 0.4963,
+      "step": 11555
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2357812910553831,
+      "learning_rate": 5.565227958056896e-07,
+      "loss": 0.4954,
+      "step": 11556
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1912208788135292,
+      "learning_rate": 5.556965535179182e-07,
+      "loss": 0.4829,
+      "step": 11557
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3335518757213265,
+      "learning_rate": 5.548709074977487e-07,
+      "loss": 0.5646,
+      "step": 11558
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3082056181625616,
+      "learning_rate": 5.540458577973118e-07,
+      "loss": 0.5054,
+      "step": 11559
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2630733151363325,
+      "learning_rate": 5.532214044686923e-07,
+      "loss": 0.4922,
+      "step": 11560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1330235251823495,
+      "learning_rate": 5.523975475639443e-07,
+      "loss": 0.4787,
+      "step": 11561
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.173797708583798,
+      "learning_rate": 5.515742871350838e-07,
+      "loss": 0.4792,
+      "step": 11562
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1836511075565557,
+      "learning_rate": 5.507516232340848e-07,
+      "loss": 0.4875,
+      "step": 11563
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.191391841613695,
+      "learning_rate": 5.499295559128892e-07,
+      "loss": 0.4711,
+      "step": 11564
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2340279586258522,
+      "learning_rate": 5.491080852233955e-07,
+      "loss": 0.5353,
+      "step": 11565
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1752932223235342,
+      "learning_rate": 5.482872112174698e-07,
+      "loss": 0.4687,
+      "step": 11566
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2523062411835026,
+      "learning_rate": 5.474669339469351e-07,
+      "loss": 0.4597,
+      "step": 11567
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.148199926183579,
+      "learning_rate": 5.466472534635814e-07,
+      "loss": 0.5053,
+      "step": 11568
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2781162084449653,
+      "learning_rate": 5.458281698191615e-07,
+      "loss": 0.5407,
+      "step": 11569
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2211590945282969,
+      "learning_rate": 5.450096830653851e-07,
+      "loss": 0.4956,
+      "step": 11570
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2045974638426897,
+      "learning_rate": 5.441917932539287e-07,
+      "loss": 0.5199,
+      "step": 11571
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1920544971354834,
+      "learning_rate": 5.433745004364299e-07,
+      "loss": 0.5059,
+      "step": 11572
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1258671385637893,
+      "learning_rate": 5.425578046644886e-07,
+      "loss": 0.4482,
+      "step": 11573
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3430994810127084,
+      "learning_rate": 5.417417059896679e-07,
+      "loss": 0.5298,
+      "step": 11574
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.30765711728383,
+      "learning_rate": 5.4092620446349e-07,
+      "loss": 0.5017,
+      "step": 11575
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3845164689572027,
+      "learning_rate": 5.401113001374459e-07,
+      "loss": 0.5462,
+      "step": 11576
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2505986687415667,
+      "learning_rate": 5.392969930629799e-07,
+      "loss": 0.4704,
+      "step": 11577
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1860635501985373,
+      "learning_rate": 5.384832832915055e-07,
+      "loss": 0.4972,
+      "step": 11578
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2304688468811966,
+      "learning_rate": 5.376701708743981e-07,
+      "loss": 0.515,
+      "step": 11579
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0690601010534004,
+      "learning_rate": 5.368576558629901e-07,
+      "loss": 0.4622,
+      "step": 11580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2907584853335778,
+      "learning_rate": 5.360457383085816e-07,
+      "loss": 0.5495,
+      "step": 11581
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3180428584168025,
+      "learning_rate": 5.352344182624336e-07,
+      "loss": 0.5582,
+      "step": 11582
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2552603662029,
+      "learning_rate": 5.344236957757687e-07,
+      "loss": 0.5372,
+      "step": 11583
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2918543781762506,
+      "learning_rate": 5.336135708997725e-07,
+      "loss": 0.5407,
+      "step": 11584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.054848778720736,
+      "learning_rate": 5.328040436855908e-07,
+      "loss": 0.453,
+      "step": 11585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.310663573389164,
+      "learning_rate": 5.31995114184336e-07,
+      "loss": 0.5397,
+      "step": 11586
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2875970544621982,
+      "learning_rate": 5.311867824470762e-07,
+      "loss": 0.5238,
+      "step": 11587
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1991838083550863,
+      "learning_rate": 5.303790485248472e-07,
+      "loss": 0.5118,
+      "step": 11588
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1703311159820398,
+      "learning_rate": 5.295719124686482e-07,
+      "loss": 0.4961,
+      "step": 11589
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1631919114320926,
+      "learning_rate": 5.28765374329433e-07,
+      "loss": 0.4735,
+      "step": 11590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2684882941722748,
+      "learning_rate": 5.279594341581252e-07,
+      "loss": 0.4885,
+      "step": 11591
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2766723734941412,
+      "learning_rate": 5.271540920056073e-07,
+      "loss": 0.5179,
+      "step": 11592
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1883534326142073,
+      "learning_rate": 5.263493479227255e-07,
+      "loss": 0.4671,
+      "step": 11593
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2319199012845141,
+      "learning_rate": 5.25545201960288e-07,
+      "loss": 0.5094,
+      "step": 11594
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2287588197191746,
+      "learning_rate": 5.247416541690642e-07,
+      "loss": 0.4652,
+      "step": 11595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.476701517978316,
+      "learning_rate": 5.239387045997835e-07,
+      "loss": 0.5068,
+      "step": 11596
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.163370989533444,
+      "learning_rate": 5.231363533031431e-07,
+      "loss": 0.4676,
+      "step": 11597
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2597525189331573,
+      "learning_rate": 5.22334600329798e-07,
+      "loss": 0.521,
+      "step": 11598
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.325676669669993,
+      "learning_rate": 5.2153344573037e-07,
+      "loss": 0.4791,
+      "step": 11599
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.256546378403438,
+      "learning_rate": 5.207328895554365e-07,
+      "loss": 0.4921,
+      "step": 11600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3392610756560481,
+      "learning_rate": 5.199329318555424e-07,
+      "loss": 0.5081,
+      "step": 11601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.20082164093368,
+      "learning_rate": 5.191335726811931e-07,
+      "loss": 0.4769,
+      "step": 11602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1291575432628187,
+      "learning_rate": 5.183348120828558e-07,
+      "loss": 0.4957,
+      "step": 11603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1954025035208389,
+      "learning_rate": 5.175366501109625e-07,
+      "loss": 0.4782,
+      "step": 11604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0798469834188902,
+      "learning_rate": 5.167390868159028e-07,
+      "loss": 0.4801,
+      "step": 11605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3001650705440604,
+      "learning_rate": 5.15942122248031e-07,
+      "loss": 0.4899,
+      "step": 11606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2203706578501075,
+      "learning_rate": 5.151457564576645e-07,
+      "loss": 0.5215,
+      "step": 11607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1625401828342996,
+      "learning_rate": 5.14349989495082e-07,
+      "loss": 0.4965,
+      "step": 11608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1946233620316558,
+      "learning_rate": 5.135548214105235e-07,
+      "loss": 0.5228,
+      "step": 11609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.171749108228106,
+      "learning_rate": 5.127602522541942e-07,
+      "loss": 0.4461,
+      "step": 11610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.240285172511022,
+      "learning_rate": 5.119662820762572e-07,
+      "loss": 0.5171,
+      "step": 11611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2563639764656076,
+      "learning_rate": 5.111729109268405e-07,
+      "loss": 0.5591,
+      "step": 11612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2266489472266957,
+      "learning_rate": 5.103801388560337e-07,
+      "loss": 0.5103,
+      "step": 11613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.097857853848195,
+      "learning_rate": 5.095879659138892e-07,
+      "loss": 0.4718,
+      "step": 11614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.306902165032515,
+      "learning_rate": 5.087963921504213e-07,
+      "loss": 0.5301,
+      "step": 11615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3801666637344339,
+      "learning_rate": 5.080054176156035e-07,
+      "loss": 0.5312,
+      "step": 11616
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1768228979828945,
+      "learning_rate": 5.072150423593769e-07,
+      "loss": 0.5405,
+      "step": 11617
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.097660051115249,
+      "learning_rate": 5.064252664316405e-07,
+      "loss": 0.4732,
+      "step": 11618
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2503430849361337,
+      "learning_rate": 5.056360898822577e-07,
+      "loss": 0.5135,
+      "step": 11619
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.16105427007145,
+      "learning_rate": 5.048475127610531e-07,
+      "loss": 0.485,
+      "step": 11620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2480716613314087,
+      "learning_rate": 5.040595351178134e-07,
+      "loss": 0.4756,
+      "step": 11621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.330302349801889,
+      "learning_rate": 5.032721570022881e-07,
+      "loss": 0.5101,
+      "step": 11622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1913059880327987,
+      "learning_rate": 5.02485378464187e-07,
+      "loss": 0.5096,
+      "step": 11623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1441361718629999,
+      "learning_rate": 5.016991995531872e-07,
+      "loss": 0.4722,
+      "step": 11624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.117943141385377,
+      "learning_rate": 5.009136203189214e-07,
+      "loss": 0.4632,
+      "step": 11625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2158038402498605,
+      "learning_rate": 5.001286408109862e-07,
+      "loss": 0.5401,
+      "step": 11626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1248110506540474,
+      "learning_rate": 4.993442610789423e-07,
+      "loss": 0.503,
+      "step": 11627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.333943356599647,
+      "learning_rate": 4.985604811723133e-07,
+      "loss": 0.5397,
+      "step": 11628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.230414882873489,
+      "learning_rate": 4.977773011405806e-07,
+      "loss": 0.4974,
+      "step": 11629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2653788103443642,
+      "learning_rate": 4.969947210331938e-07,
+      "loss": 0.542,
+      "step": 11630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.094276301597115,
+      "learning_rate": 4.962127408995587e-07,
+      "loss": 0.4612,
+      "step": 11631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1960067727366028,
+      "learning_rate": 4.95431360789046e-07,
+      "loss": 0.469,
+      "step": 11632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.20365740403089,
+      "learning_rate": 4.946505807509883e-07,
+      "loss": 0.487,
+      "step": 11633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2335867960699443,
+      "learning_rate": 4.938704008346818e-07,
+      "loss": 0.5124,
+      "step": 11634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2364535155144885,
+      "learning_rate": 4.930908210893826e-07,
+      "loss": 0.4667,
+      "step": 11635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3809632258232114,
+      "learning_rate": 4.92311841564308e-07,
+      "loss": 0.562,
+      "step": 11636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1480338496624145,
+      "learning_rate": 4.915334623086387e-07,
+      "loss": 0.4954,
+      "step": 11637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.172027832237827,
+      "learning_rate": 4.907556833715199e-07,
+      "loss": 0.4747,
+      "step": 11638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2427988047680911,
+      "learning_rate": 4.899785048020567e-07,
+      "loss": 0.4948,
+      "step": 11639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.170583241904792,
+      "learning_rate": 4.892019266493164e-07,
+      "loss": 0.5279,
+      "step": 11640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.196469314385757,
+      "learning_rate": 4.884259489623267e-07,
+      "loss": 0.475,
+      "step": 11641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1514619718691659,
+      "learning_rate": 4.876505717900803e-07,
+      "loss": 0.4724,
+      "step": 11642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2511098226403587,
+      "learning_rate": 4.868757951815295e-07,
+      "loss": 0.5248,
+      "step": 11643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3178662091193285,
+      "learning_rate": 4.861016191855939e-07,
+      "loss": 0.5234,
+      "step": 11644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2354141400436287,
+      "learning_rate": 4.853280438511476e-07,
+      "loss": 0.5092,
+      "step": 11645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.098511440778354,
+      "learning_rate": 4.845550692270296e-07,
+      "loss": 0.498,
+      "step": 11646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2128744078210814,
+      "learning_rate": 4.837826953620428e-07,
+      "loss": 0.4699,
+      "step": 11647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1940351654893464,
+      "learning_rate": 4.830109223049528e-07,
+      "loss": 0.4796,
+      "step": 11648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2091029883946196,
+      "learning_rate": 4.82239750104484e-07,
+      "loss": 0.5288,
+      "step": 11649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1562578871174463,
+      "learning_rate": 4.814691788093262e-07,
+      "loss": 0.4894,
+      "step": 11650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3747158190356996,
+      "learning_rate": 4.806992084681273e-07,
+      "loss": 0.5272,
+      "step": 11651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2156153739803044,
+      "learning_rate": 4.799298391295004e-07,
+      "loss": 0.5203,
+      "step": 11652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1915433210724786,
+      "learning_rate": 4.791610708420192e-07,
+      "loss": 0.4939,
+      "step": 11653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3170288605826161,
+      "learning_rate": 4.783929036542234e-07,
+      "loss": 0.5232,
+      "step": 11654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2679209664320248,
+      "learning_rate": 4.776253376146078e-07,
+      "loss": 0.5041,
+      "step": 11655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1599221041438859,
+      "learning_rate": 4.768583727716314e-07,
+      "loss": 0.4681,
+      "step": 11656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1336253965601941,
+      "learning_rate": 4.760920091737198e-07,
+      "loss": 0.473,
+      "step": 11657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1305282186962131,
+      "learning_rate": 4.7532624686925655e-07,
+      "loss": 0.4607,
+      "step": 11658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1431852374257534,
+      "learning_rate": 4.745610859065886e-07,
+      "loss": 0.4501,
+      "step": 11659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3306119634546028,
+      "learning_rate": 4.7379652633402386e-07,
+      "loss": 0.4928,
+      "step": 11660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1978248173980799,
+      "learning_rate": 4.730325681998338e-07,
+      "loss": 0.5595,
+      "step": 11661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.219627015913165,
+      "learning_rate": 4.722692115522498e-07,
+      "loss": 0.5345,
+      "step": 11662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.223947982203224,
+      "learning_rate": 4.715064564394667e-07,
+      "loss": 0.5003,
+      "step": 11663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2324123472882256,
+      "learning_rate": 4.707443029096437e-07,
+      "loss": 0.457,
+      "step": 11664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0759650445315745,
+      "learning_rate": 4.6998275101089454e-07,
+      "loss": 0.4797,
+      "step": 11665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.185037468909366,
+      "learning_rate": 4.692218007913052e-07,
+      "loss": 0.5174,
+      "step": 11666
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2233893783284107,
+      "learning_rate": 4.6846145229891393e-07,
+      "loss": 0.4962,
+      "step": 11667
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2070307561879552,
+      "learning_rate": 4.677017055817279e-07,
+      "loss": 0.498,
+      "step": 11668
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.306967017401899,
+      "learning_rate": 4.6694256068771314e-07,
+      "loss": 0.5078,
+      "step": 11669
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1385053509606156,
+      "learning_rate": 4.661840176647991e-07,
+      "loss": 0.4506,
+      "step": 11670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.260228981146994,
+      "learning_rate": 4.654260765608776e-07,
+      "loss": 0.5134,
+      "step": 11671
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2414791558705525,
+      "learning_rate": 4.646687374237979e-07,
+      "loss": 0.481,
+      "step": 11672
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.224950569479002,
+      "learning_rate": 4.639120003013764e-07,
+      "loss": 0.4756,
+      "step": 11673
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2187451093526673,
+      "learning_rate": 4.6315586524139143e-07,
+      "loss": 0.537,
+      "step": 11674
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3131997649938285,
+      "learning_rate": 4.6240033229157934e-07,
+      "loss": 0.563,
+      "step": 11675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1431704819663746,
+      "learning_rate": 4.61645401499643e-07,
+      "loss": 0.4919,
+      "step": 11676
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1505580190943998,
+      "learning_rate": 4.608910729132432e-07,
+      "loss": 0.4722,
+      "step": 11677
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1179001142651064,
+      "learning_rate": 4.6013734658000406e-07,
+      "loss": 0.4819,
+      "step": 11678
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2458995799352273,
+      "learning_rate": 4.5938422254751425e-07,
+      "loss": 0.5234,
+      "step": 11679
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1915886411042884,
+      "learning_rate": 4.5863170086332234e-07,
+      "loss": 0.5037,
+      "step": 11680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1996697189869534,
+      "learning_rate": 4.578797815749381e-07,
+      "loss": 0.4958,
+      "step": 11681
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3254755858047529,
+      "learning_rate": 4.571284647298335e-07,
+      "loss": 0.5179,
+      "step": 11682
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4197493373562327,
+      "learning_rate": 4.5637775037544296e-07,
+      "loss": 0.5397,
+      "step": 11683
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1658080665480766,
+      "learning_rate": 4.556276385591663e-07,
+      "loss": 0.4958,
+      "step": 11684
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.196314472784978,
+      "learning_rate": 4.548781293283566e-07,
+      "loss": 0.4759,
+      "step": 11685
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.084411152173807,
+      "learning_rate": 4.541292227303384e-07,
+      "loss": 0.4522,
+      "step": 11686
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.252107131697988,
+      "learning_rate": 4.533809188123917e-07,
+      "loss": 0.5192,
+      "step": 11687
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.241396430176583,
+      "learning_rate": 4.526332176217618e-07,
+      "loss": 0.5326,
+      "step": 11688
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1364092848018812,
+      "learning_rate": 4.518861192056545e-07,
+      "loss": 0.4621,
+      "step": 11689
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3141722697926925,
+      "learning_rate": 4.511396236112386e-07,
+      "loss": 0.4955,
+      "step": 11690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2908912864286366,
+      "learning_rate": 4.503937308856454e-07,
+      "loss": 0.5272,
+      "step": 11691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1712128866823428,
+      "learning_rate": 4.4964844107596384e-07,
+      "loss": 0.4973,
+      "step": 11692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1982866572776498,
+      "learning_rate": 4.489037542292507e-07,
+      "loss": 0.4444,
+      "step": 11693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.172101164292414,
+      "learning_rate": 4.481596703925195e-07,
+      "loss": 0.4915,
+      "step": 11694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2125661419004228,
+      "learning_rate": 4.4741618961274936e-07,
+      "loss": 0.4956,
+      "step": 11695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.153808335454769,
+      "learning_rate": 4.4667331193688155e-07,
+      "loss": 0.4789,
+      "step": 11696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1952670718268237,
+      "learning_rate": 4.459310374118142e-07,
+      "loss": 0.4465,
+      "step": 11697
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3161940587573808,
+      "learning_rate": 4.451893660844142e-07,
+      "loss": 0.4697,
+      "step": 11698
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2213759376304432,
+      "learning_rate": 4.4444829800150524e-07,
+      "loss": 0.4709,
+      "step": 11699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.192370315473772,
+      "learning_rate": 4.437078332098754e-07,
+      "loss": 0.4807,
+      "step": 11700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2790178067367342,
+      "learning_rate": 4.4296797175627517e-07,
+      "loss": 0.5051,
+      "step": 11701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.199433347903852,
+      "learning_rate": 4.422287136874126e-07,
+      "loss": 0.472,
+      "step": 11702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2479101353868722,
+      "learning_rate": 4.41490059049966e-07,
+      "loss": 0.5427,
+      "step": 11703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.159155005792415,
+      "learning_rate": 4.407520078905647e-07,
+      "loss": 0.4453,
+      "step": 11704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.225054402992305,
+      "learning_rate": 4.4001456025580925e-07,
+      "loss": 0.5116,
+      "step": 11705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1330927535032986,
+      "learning_rate": 4.3927771619225787e-07,
+      "loss": 0.4546,
+      "step": 11706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3007622794752653,
+      "learning_rate": 4.385414757464312e-07,
+      "loss": 0.4898,
+      "step": 11707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2646846810232406,
+      "learning_rate": 4.37805838964811e-07,
+      "loss": 0.5187,
+      "step": 11708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2703951669143438,
+      "learning_rate": 4.370708058938422e-07,
+      "loss": 0.5491,
+      "step": 11709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1870486757024492,
+      "learning_rate": 4.363363765799322e-07,
+      "loss": 0.5152,
+      "step": 11710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.098214167197262,
+      "learning_rate": 4.356025510694495e-07,
+      "loss": 0.5115,
+      "step": 11711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2990648555848965,
+      "learning_rate": 4.348693294087236e-07,
+      "loss": 0.5695,
+      "step": 11712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1769447526402916,
+      "learning_rate": 4.3413671164404757e-07,
+      "loss": 0.5,
+      "step": 11713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2447782646106818,
+      "learning_rate": 4.3340469782167214e-07,
+      "loss": 0.476,
+      "step": 11714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2250338705401749,
+      "learning_rate": 4.3267328798781595e-07,
+      "loss": 0.5429,
+      "step": 11715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3735424466012531,
+      "learning_rate": 4.319424821886553e-07,
+      "loss": 0.5384,
+      "step": 11716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1890178817418209,
+      "learning_rate": 4.3121228047033227e-07,
+      "loss": 0.4781,
+      "step": 11717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.157881384896909,
+      "learning_rate": 4.3048268287894435e-07,
+      "loss": 0.4957,
+      "step": 11718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2604530056005232,
+      "learning_rate": 4.2975368946055805e-07,
+      "loss": 0.4931,
+      "step": 11719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2466276932215556,
+      "learning_rate": 4.290253002611966e-07,
+      "loss": 0.4998,
+      "step": 11720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1932600763591998,
+      "learning_rate": 4.282975153268476e-07,
+      "loss": 0.4837,
+      "step": 11721
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2554693254090885,
+      "learning_rate": 4.275703347034621e-07,
+      "loss": 0.558,
+      "step": 11722
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3863700750715937,
+      "learning_rate": 4.268437584369478e-07,
+      "loss": 0.5683,
+      "step": 11723
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2086752648904282,
+      "learning_rate": 4.2611778657317695e-07,
+      "loss": 0.5212,
+      "step": 11724
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1653620998704188,
+      "learning_rate": 4.25392419157985e-07,
+      "loss": 0.4999,
+      "step": 11725
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2040753265022415,
+      "learning_rate": 4.2466765623716766e-07,
+      "loss": 0.5015,
+      "step": 11726
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3222846927141374,
+      "learning_rate": 4.2394349785648494e-07,
+      "loss": 0.5322,
+      "step": 11727
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2089381293848507,
+      "learning_rate": 4.232199440616536e-07,
+      "loss": 0.4804,
+      "step": 11728
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1606385740802105,
+      "learning_rate": 4.2249699489835815e-07,
+      "loss": 0.4857,
+      "step": 11729
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3053066589064044,
+      "learning_rate": 4.2177465041223995e-07,
+      "loss": 0.5312,
+      "step": 11730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1614812612052423,
+      "learning_rate": 4.2105291064890474e-07,
+      "loss": 0.4595,
+      "step": 11731
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1995305951260657,
+      "learning_rate": 4.203317756539216e-07,
+      "loss": 0.5121,
+      "step": 11732
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2239049805179607,
+      "learning_rate": 4.196112454728185e-07,
+      "loss": 0.5229,
+      "step": 11733
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2171196915000573,
+      "learning_rate": 4.1889132015108467e-07,
+      "loss": 0.5001,
+      "step": 11734
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8535005751743459,
+      "learning_rate": 4.181719997341738e-07,
+      "loss": 0.4794,
+      "step": 11735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2263811791392332,
+      "learning_rate": 4.1745328426749943e-07,
+      "loss": 0.5545,
+      "step": 11736
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2140691317618955,
+      "learning_rate": 4.167351737964409e-07,
+      "loss": 0.5407,
+      "step": 11737
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.247686151899819,
+      "learning_rate": 4.1601766836633196e-07,
+      "loss": 0.5128,
+      "step": 11738
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1683219452183686,
+      "learning_rate": 4.153007680224752e-07,
+      "loss": 0.5228,
+      "step": 11739
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2837260906515053,
+      "learning_rate": 4.1458447281013113e-07,
+      "loss": 0.4502,
+      "step": 11740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2071021593122753,
+      "learning_rate": 4.138687827745236e-07,
+      "loss": 0.4725,
+      "step": 11741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2537044469710232,
+      "learning_rate": 4.131536979608386e-07,
+      "loss": 0.4991,
+      "step": 11742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.193695520486006,
+      "learning_rate": 4.124392184142223e-07,
+      "loss": 0.4982,
+      "step": 11743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1998289959975694,
+      "learning_rate": 4.1172534417978305e-07,
+      "loss": 0.5043,
+      "step": 11744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1520581424895415,
+      "learning_rate": 4.1101207530259144e-07,
+      "loss": 0.4442,
+      "step": 11745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1643199763911125,
+      "learning_rate": 4.102994118276804e-07,
+      "loss": 0.5122,
+      "step": 11746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1934336437833286,
+      "learning_rate": 4.095873538000439e-07,
+      "loss": 0.5159,
+      "step": 11747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.253856051857348,
+      "learning_rate": 4.088759012646382e-07,
+      "loss": 0.486,
+      "step": 11748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1820389005485281,
+      "learning_rate": 4.081650542663795e-07,
+      "loss": 0.5113,
+      "step": 11749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1781774797564402,
+      "learning_rate": 4.0745481285014876e-07,
+      "loss": 0.483,
+      "step": 11750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0738288310448638,
+      "learning_rate": 4.0674517706078556e-07,
+      "loss": 0.4665,
+      "step": 11751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2804562971290436,
+      "learning_rate": 4.060361469430962e-07,
+      "loss": 0.503,
+      "step": 11752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2892818379919415,
+      "learning_rate": 4.0532772254184394e-07,
+      "loss": 0.5132,
+      "step": 11753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2004887817252679,
+      "learning_rate": 4.0461990390175175e-07,
+      "loss": 0.4391,
+      "step": 11754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2543431646833745,
+      "learning_rate": 4.0391269106751174e-07,
+      "loss": 0.5051,
+      "step": 11755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2769473799526383,
+      "learning_rate": 4.032060840837726e-07,
+      "loss": 0.4994,
+      "step": 11756
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1998899985599114,
+      "learning_rate": 4.0250008299514755e-07,
+      "loss": 0.4979,
+      "step": 11757
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2316834766254283,
+      "learning_rate": 4.017946878462076e-07,
+      "loss": 0.4457,
+      "step": 11758
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1375048039932254,
+      "learning_rate": 4.010898986814893e-07,
+      "loss": 0.4964,
+      "step": 11759
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.299131383842934,
+      "learning_rate": 4.0038571554548934e-07,
+      "loss": 0.5086,
+      "step": 11760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1954543584282178,
+      "learning_rate": 3.9968213848266655e-07,
+      "loss": 0.4802,
+      "step": 11761
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2054921432865644,
+      "learning_rate": 3.9897916753744324e-07,
+      "loss": 0.5338,
+      "step": 11762
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2764553048407306,
+      "learning_rate": 3.9827680275419944e-07,
+      "loss": 0.5121,
+      "step": 11763
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2053758942694601,
+      "learning_rate": 3.9757504417727856e-07,
+      "loss": 0.4842,
+      "step": 11764
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1320032846854713,
+      "learning_rate": 3.9687389185098733e-07,
+      "loss": 0.4689,
+      "step": 11765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.214284639398115,
+      "learning_rate": 3.9617334581959267e-07,
+      "loss": 0.5429,
+      "step": 11766
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2946557969823145,
+      "learning_rate": 3.9547340612732356e-07,
+      "loss": 0.5099,
+      "step": 11767
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2584743298658398,
+      "learning_rate": 3.9477407281837246e-07,
+      "loss": 0.5073,
+      "step": 11768
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1653233810330084,
+      "learning_rate": 3.940753459368896e-07,
+      "loss": 0.4227,
+      "step": 11769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2595806133826835,
+      "learning_rate": 3.9337722552698963e-07,
+      "loss": 0.5476,
+      "step": 11770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2531475968832562,
+      "learning_rate": 3.9267971163274966e-07,
+      "loss": 0.5068,
+      "step": 11771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0947538401381067,
+      "learning_rate": 3.919828042982077e-07,
+      "loss": 0.45,
+      "step": 11772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2633416571296776,
+      "learning_rate": 3.9128650356736297e-07,
+      "loss": 0.4809,
+      "step": 11773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.237929333279952,
+      "learning_rate": 3.905908094841737e-07,
+      "loss": 0.5068,
+      "step": 11774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1030452299588724,
+      "learning_rate": 3.898957220925648e-07,
+      "loss": 0.4438,
+      "step": 11775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2417517803556468,
+      "learning_rate": 3.8920124143642104e-07,
+      "loss": 0.5079,
+      "step": 11776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1804708444209424,
+      "learning_rate": 3.885073675595874e-07,
+      "loss": 0.4584,
+      "step": 11777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0939406092630664,
+      "learning_rate": 3.878141005058733e-07,
+      "loss": 0.454,
+      "step": 11778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1600049880348036,
+      "learning_rate": 3.871214403190471e-07,
+      "loss": 0.4961,
+      "step": 11779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2780535297883957,
+      "learning_rate": 3.864293870428404e-07,
+      "loss": 0.5328,
+      "step": 11780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2095696876606805,
+      "learning_rate": 3.8573794072094495e-07,
+      "loss": 0.5176,
+      "step": 11781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2659270785544752,
+      "learning_rate": 3.8504710139701804e-07,
+      "loss": 0.5117,
+      "step": 11782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.319515461711487,
+      "learning_rate": 3.843568691146748e-07,
+      "loss": 0.5153,
+      "step": 11783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2002694562539997,
+      "learning_rate": 3.8366724391749153e-07,
+      "loss": 0.5164,
+      "step": 11784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2488768299903783,
+      "learning_rate": 3.829782258490078e-07,
+      "loss": 0.4832,
+      "step": 11785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3546710126636907,
+      "learning_rate": 3.8228981495272654e-07,
+      "loss": 0.5357,
+      "step": 11786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2989807039285735,
+      "learning_rate": 3.8160201127211083e-07,
+      "loss": 0.489,
+      "step": 11787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1660209924008844,
+      "learning_rate": 3.809148148505848e-07,
+      "loss": 0.477,
+      "step": 11788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2577406110418798,
+      "learning_rate": 3.8022822573153374e-07,
+      "loss": 0.4625,
+      "step": 11789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.103991866520623,
+      "learning_rate": 3.795422439583063e-07,
+      "loss": 0.4281,
+      "step": 11790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1095122131596897,
+      "learning_rate": 3.788568695742123e-07,
+      "loss": 0.4694,
+      "step": 11791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2214688028356104,
+      "learning_rate": 3.781721026225227e-07,
+      "loss": 0.5078,
+      "step": 11792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2617199310202725,
+      "learning_rate": 3.7748794314647066e-07,
+      "loss": 0.5729,
+      "step": 11793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2734457378472044,
+      "learning_rate": 3.768043911892505e-07,
+      "loss": 0.539,
+      "step": 11794
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.40469694208657,
+      "learning_rate": 3.7612144679401664e-07,
+      "loss": 0.5234,
+      "step": 11795
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1789927679006709,
+      "learning_rate": 3.754391100038901e-07,
+      "loss": 0.5356,
+      "step": 11796
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2780748427419666,
+      "learning_rate": 3.747573808619476e-07,
+      "loss": 0.5241,
+      "step": 11797
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2062431651499914,
+      "learning_rate": 3.740762594112324e-07,
+      "loss": 0.4773,
+      "step": 11798
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.195164340941025,
+      "learning_rate": 3.733957456947457e-07,
+      "loss": 0.5391,
+      "step": 11799
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1602979888131897,
+      "learning_rate": 3.72715839755452e-07,
+      "loss": 0.4759,
+      "step": 11800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2576929355445683,
+      "learning_rate": 3.720365416362792e-07,
+      "loss": 0.5166,
+      "step": 11801
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.248847239147885,
+      "learning_rate": 3.713578513801119e-07,
+      "loss": 0.5368,
+      "step": 11802
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2689911148145008,
+      "learning_rate": 3.706797690298014e-07,
+      "loss": 0.5151,
+      "step": 11803
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1390971049009015,
+      "learning_rate": 3.7000229462815785e-07,
+      "loss": 0.4449,
+      "step": 11804
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2716373287148128,
+      "learning_rate": 3.6932542821795256e-07,
+      "loss": 0.5376,
+      "step": 11805
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2276580194511737,
+      "learning_rate": 3.6864916984192143e-07,
+      "loss": 0.4925,
+      "step": 11806
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.201898469819754,
+      "learning_rate": 3.6797351954275916e-07,
+      "loss": 0.4607,
+      "step": 11807
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2896428044687338,
+      "learning_rate": 3.6729847736312387e-07,
+      "loss": 0.5385,
+      "step": 11808
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2846397127112779,
+      "learning_rate": 3.6662404334563363e-07,
+      "loss": 0.5583,
+      "step": 11809
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2214329361339653,
+      "learning_rate": 3.6595021753286886e-07,
+      "loss": 0.4906,
+      "step": 11810
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2725794423005357,
+      "learning_rate": 3.652769999673733e-07,
+      "loss": 0.5469,
+      "step": 11811
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2037241546137951,
+      "learning_rate": 3.646043906916474e-07,
+      "loss": 0.5198,
+      "step": 11812
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.360362297284121,
+      "learning_rate": 3.639323897481606e-07,
+      "loss": 0.5368,
+      "step": 11813
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2828800717955788,
+      "learning_rate": 3.632609971793366e-07,
+      "loss": 0.4928,
+      "step": 11814
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.11524998107787,
+      "learning_rate": 3.6259021302756383e-07,
+      "loss": 0.4862,
+      "step": 11815
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1956216250350546,
+      "learning_rate": 3.619200373351939e-07,
+      "loss": 0.4736,
+      "step": 11816
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2944798240047037,
+      "learning_rate": 3.612504701445385e-07,
+      "loss": 0.4855,
+      "step": 11817
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.192429400259517,
+      "learning_rate": 3.6058151149787166e-07,
+      "loss": 0.5089,
+      "step": 11818
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.350053683732373,
+      "learning_rate": 3.5991316143742515e-07,
+      "loss": 0.4956,
+      "step": 11819
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2975610504046746,
+      "learning_rate": 3.592454200053963e-07,
+      "loss": 0.5721,
+      "step": 11820
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2104513853853138,
+      "learning_rate": 3.585782872439458e-07,
+      "loss": 0.5139,
+      "step": 11821
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.318205964381494,
+      "learning_rate": 3.5791176319519006e-07,
+      "loss": 0.5689,
+      "step": 11822
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.205212108020372,
+      "learning_rate": 3.5724584790121084e-07,
+      "loss": 0.4791,
+      "step": 11823
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2967040627210051,
+      "learning_rate": 3.565805414040535e-07,
+      "loss": 0.5225,
+      "step": 11824
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.326910742890564,
+      "learning_rate": 3.5591584374571773e-07,
+      "loss": 0.5302,
+      "step": 11825
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1723913962326695,
+      "learning_rate": 3.5525175496817223e-07,
+      "loss": 0.4868,
+      "step": 11826
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1615463303330051,
+      "learning_rate": 3.545882751133445e-07,
+      "loss": 0.502,
+      "step": 11827
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.225542652839979,
+      "learning_rate": 3.5392540422312213e-07,
+      "loss": 0.4472,
+      "step": 11828
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1211358570870824,
+      "learning_rate": 3.5326314233935734e-07,
+      "loss": 0.5319,
+      "step": 11829
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2231918478092207,
+      "learning_rate": 3.5260148950385985e-07,
+      "loss": 0.4812,
+      "step": 11830
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.208629944449851,
+      "learning_rate": 3.5194044575840523e-07,
+      "loss": 0.4487,
+      "step": 11831
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.052311686796946,
+      "learning_rate": 3.5128001114472674e-07,
+      "loss": 0.4481,
+      "step": 11832
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2427054710813539,
+      "learning_rate": 3.506201857045222e-07,
+      "loss": 0.4825,
+      "step": 11833
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1732240668177534,
+      "learning_rate": 3.4996096947945036e-07,
+      "loss": 0.4688,
+      "step": 11834
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.25371243413933,
+      "learning_rate": 3.4930236251112914e-07,
+      "loss": 0.4958,
+      "step": 11835
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1516007436460405,
+      "learning_rate": 3.4864436484114086e-07,
+      "loss": 0.4779,
+      "step": 11836
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2052656179324965,
+      "learning_rate": 3.4798697651102887e-07,
+      "loss": 0.5564,
+      "step": 11837
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2081124608099083,
+      "learning_rate": 3.4733019756229557e-07,
+      "loss": 0.4667,
+      "step": 11838
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.194664025022433,
+      "learning_rate": 3.4667402803641005e-07,
+      "loss": 0.4947,
+      "step": 11839
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0570926510274017,
+      "learning_rate": 3.460184679747969e-07,
+      "loss": 0.4169,
+      "step": 11840
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.119244316050855,
+      "learning_rate": 3.453635174188463e-07,
+      "loss": 0.4697,
+      "step": 11841
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2132183129861274,
+      "learning_rate": 3.447091764099075e-07,
+      "loss": 0.4992,
+      "step": 11842
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2914518003282607,
+      "learning_rate": 3.440554449892941e-07,
+      "loss": 0.5004,
+      "step": 11843
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1518716653329535,
+      "learning_rate": 3.434023231982786e-07,
+      "loss": 0.4562,
+      "step": 11844
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2147686312593688,
+      "learning_rate": 3.4274981107809466e-07,
+      "loss": 0.4195,
+      "step": 11845
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1839667509926344,
+      "learning_rate": 3.4209790866994055e-07,
+      "loss": 0.4858,
+      "step": 11846
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.171469554699259,
+      "learning_rate": 3.414466160149732e-07,
+      "loss": 0.4887,
+      "step": 11847
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.210771660062191,
+      "learning_rate": 3.4079593315431315e-07,
+      "loss": 0.4957,
+      "step": 11848
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2522919147351381,
+      "learning_rate": 3.401458601290408e-07,
+      "loss": 0.534,
+      "step": 11849
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2936390409939664,
+      "learning_rate": 3.3949639698019896e-07,
+      "loss": 0.5641,
+      "step": 11850
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.369830731703924,
+      "learning_rate": 3.388475437487915e-07,
+      "loss": 0.5514,
+      "step": 11851
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.297526230519369,
+      "learning_rate": 3.381993004757822e-07,
+      "loss": 0.5662,
+      "step": 11852
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2452877870259023,
+      "learning_rate": 3.3755166720210065e-07,
+      "loss": 0.5143,
+      "step": 11853
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1829554785076084,
+      "learning_rate": 3.3690464396863407e-07,
+      "loss": 0.5026,
+      "step": 11854
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1929361994711225,
+      "learning_rate": 3.3625823081623097e-07,
+      "loss": 0.4908,
+      "step": 11855
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.153294213344485,
+      "learning_rate": 3.3561242778570426e-07,
+      "loss": 0.5194,
+      "step": 11856
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2180884106123477,
+      "learning_rate": 3.349672349178279e-07,
+      "loss": 0.466,
+      "step": 11857
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2845498367447854,
+      "learning_rate": 3.343226522533338e-07,
+      "loss": 0.558,
+      "step": 11858
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.172581777425249,
+      "learning_rate": 3.3367867983292056e-07,
+      "loss": 0.4776,
+      "step": 11859
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.239165414622704,
+      "learning_rate": 3.330353176972423e-07,
+      "loss": 0.47,
+      "step": 11860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1473822384082426,
+      "learning_rate": 3.323925658869209e-07,
+      "loss": 0.4775,
+      "step": 11861
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1340337024437142,
+      "learning_rate": 3.3175042444253405e-07,
+      "loss": 0.4497,
+      "step": 11862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2249561652300989,
+      "learning_rate": 3.311088934046247e-07,
+      "loss": 0.5151,
+      "step": 11863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1540089617700202,
+      "learning_rate": 3.3046797281369614e-07,
+      "loss": 0.4877,
+      "step": 11864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2289523999375467,
+      "learning_rate": 3.298276627102126e-07,
+      "loss": 0.4859,
+      "step": 11865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.187750940411227,
+      "learning_rate": 3.2918796313459954e-07,
+      "loss": 0.512,
+      "step": 11866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.19838081434691,
+      "learning_rate": 3.285488741272458e-07,
+      "loss": 0.4784,
+      "step": 11867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2647895881502962,
+      "learning_rate": 3.2791039572849903e-07,
+      "loss": 0.4906,
+      "step": 11868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2120988260806527,
+      "learning_rate": 3.272725279786715e-07,
+      "loss": 0.5222,
+      "step": 11869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1655362931520417,
+      "learning_rate": 3.2663527091803317e-07,
+      "loss": 0.4923,
+      "step": 11870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.300798891423282,
+      "learning_rate": 3.2599862458681963e-07,
+      "loss": 0.5189,
+      "step": 11871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2424211582214633,
+      "learning_rate": 3.2536258902522323e-07,
+      "loss": 0.507,
+      "step": 11872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2824541341748668,
+      "learning_rate": 3.247271642734007e-07,
+      "loss": 0.4833,
+      "step": 11873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2379879769648892,
+      "learning_rate": 3.24092350371471e-07,
+      "loss": 0.5265,
+      "step": 11874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.242445577047439,
+      "learning_rate": 3.234581473595122e-07,
+      "loss": 0.4934,
+      "step": 11875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2357918056645432,
+      "learning_rate": 3.228245552775633e-07,
+      "loss": 0.4952,
+      "step": 11876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1593829316641642,
+      "learning_rate": 3.22191574165629e-07,
+      "loss": 0.5036,
+      "step": 11877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.242463950818148,
+      "learning_rate": 3.215592040636717e-07,
+      "loss": 0.5641,
+      "step": 11878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2007545304412486,
+      "learning_rate": 3.209274450116162e-07,
+      "loss": 0.5184,
+      "step": 11879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2131656451641326,
+      "learning_rate": 3.202962970493484e-07,
+      "loss": 0.5185,
+      "step": 11880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2582918284478506,
+      "learning_rate": 3.196657602167175e-07,
+      "loss": 0.4557,
+      "step": 11881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2227858209660794,
+      "learning_rate": 3.1903583455352937e-07,
+      "loss": 0.5374,
+      "step": 11882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.168920531046059,
+      "learning_rate": 3.1840652009955563e-07,
+      "loss": 0.4883,
+      "step": 11883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3303419571299668,
+      "learning_rate": 3.1777781689453e-07,
+      "loss": 0.5477,
+      "step": 11884
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2804068140175722,
+      "learning_rate": 3.1714972497814413e-07,
+      "loss": 0.5048,
+      "step": 11885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1467945517615132,
+      "learning_rate": 3.1652224439005287e-07,
+      "loss": 0.4773,
+      "step": 11886
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.174258846529829,
+      "learning_rate": 3.158953751698723e-07,
+      "loss": 0.4696,
+      "step": 11887
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.260874510069792,
+      "learning_rate": 3.152691173571809e-07,
+      "loss": 0.5119,
+      "step": 11888
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.139834299760827,
+      "learning_rate": 3.146434709915158e-07,
+      "loss": 0.481,
+      "step": 11889
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1358686591569918,
+      "learning_rate": 3.1401843611237993e-07,
+      "loss": 0.477,
+      "step": 11890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.268411042359718,
+      "learning_rate": 3.1339401275923277e-07,
+      "loss": 0.4924,
+      "step": 11891
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2053667461678765,
+      "learning_rate": 3.127702009714961e-07,
+      "loss": 0.49,
+      "step": 11892
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.159756524168175,
+      "learning_rate": 3.121470007885574e-07,
+      "loss": 0.4644,
+      "step": 11893
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2298578108035971,
+      "learning_rate": 3.1152441224976073e-07,
+      "loss": 0.4833,
+      "step": 11894
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2799199807218755,
+      "learning_rate": 3.1090243539441565e-07,
+      "loss": 0.5611,
+      "step": 11895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.31700301863861,
+      "learning_rate": 3.1028107026178756e-07,
+      "loss": 0.5299,
+      "step": 11896
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1041180881874268,
+      "learning_rate": 3.096603168911072e-07,
+      "loss": 0.4529,
+      "step": 11897
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2636458852651218,
+      "learning_rate": 3.090401753215677e-07,
+      "loss": 0.5014,
+      "step": 11898
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3141831096587264,
+      "learning_rate": 3.084206455923211e-07,
+      "loss": 0.5246,
+      "step": 11899
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1742321467951986,
+      "learning_rate": 3.078017277424805e-07,
+      "loss": 0.5124,
+      "step": 11900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2465338811603952,
+      "learning_rate": 3.071834218111225e-07,
+      "loss": 0.5203,
+      "step": 11901
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2592303891134542,
+      "learning_rate": 3.0656572783728247e-07,
+      "loss": 0.5941,
+      "step": 11902
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1550183308785742,
+      "learning_rate": 3.059486458599592e-07,
+      "loss": 0.4972,
+      "step": 11903
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3015029325741627,
+      "learning_rate": 3.0533217591811383e-07,
+      "loss": 0.5139,
+      "step": 11904
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.267953261717419,
+      "learning_rate": 3.0471631805066626e-07,
+      "loss": 0.4956,
+      "step": 11905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.265810729162535,
+      "learning_rate": 3.0410107229649764e-07,
+      "loss": 0.4983,
+      "step": 11906
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2888394711921056,
+      "learning_rate": 3.034864386944525e-07,
+      "loss": 0.5086,
+      "step": 11907
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.255999425313903,
+      "learning_rate": 3.028724172833364e-07,
+      "loss": 0.5289,
+      "step": 11908
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2235600844900756,
+      "learning_rate": 3.022590081019161e-07,
+      "loss": 0.5266,
+      "step": 11909
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3734475388230556,
+      "learning_rate": 3.0164621118891733e-07,
+      "loss": 0.4774,
+      "step": 11910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1895065420900075,
+      "learning_rate": 3.010340265830314e-07,
+      "loss": 0.5026,
+      "step": 11911
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.233633180631941,
+      "learning_rate": 3.0042245432290616e-07,
+      "loss": 0.5101,
+      "step": 11912
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2901873680188227,
+      "learning_rate": 2.998114944471542e-07,
+      "loss": 0.5005,
+      "step": 11913
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.232769948710911,
+      "learning_rate": 2.992011469943501e-07,
+      "loss": 0.518,
+      "step": 11914
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1740572637502646,
+      "learning_rate": 2.985914120030275e-07,
+      "loss": 0.4807,
+      "step": 11915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1624653247050696,
+      "learning_rate": 2.979822895116802e-07,
+      "loss": 0.465,
+      "step": 11916
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1991851503716289,
+      "learning_rate": 2.973737795587672e-07,
+      "loss": 0.4782,
+      "step": 11917
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0942294568544784,
+      "learning_rate": 2.967658821827069e-07,
+      "loss": 0.4484,
+      "step": 11918
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2111436237750206,
+      "learning_rate": 2.9615859742187944e-07,
+      "loss": 0.5323,
+      "step": 11919
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4270931531285191,
+      "learning_rate": 2.955519253146233e-07,
+      "loss": 0.5519,
+      "step": 11920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.215772610996343,
+      "learning_rate": 2.9494586589924434e-07,
+      "loss": 0.5311,
+      "step": 11921
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2044263960002966,
+      "learning_rate": 2.9434041921400204e-07,
+      "loss": 0.5183,
+      "step": 11922
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1510258784297343,
+      "learning_rate": 2.9373558529712466e-07,
+      "loss": 0.4968,
+      "step": 11923
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.268845921916506,
+      "learning_rate": 2.9313136418679835e-07,
+      "loss": 0.4591,
+      "step": 11924
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1582245848856378,
+      "learning_rate": 2.9252775592116924e-07,
+      "loss": 0.4784,
+      "step": 11925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2377338342629929,
+      "learning_rate": 2.919247605383468e-07,
+      "loss": 0.5133,
+      "step": 11926
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3280985212491447,
+      "learning_rate": 2.913223780764007e-07,
+      "loss": 0.5569,
+      "step": 11927
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0928988278254506,
+      "learning_rate": 2.9072060857336384e-07,
+      "loss": 0.4218,
+      "step": 11928
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.110030558139715,
+      "learning_rate": 2.9011945206722904e-07,
+      "loss": 0.4884,
+      "step": 11929
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.134434347990986,
+      "learning_rate": 2.895189085959482e-07,
+      "loss": 0.4737,
+      "step": 11930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1509882308961452,
+      "learning_rate": 2.8891897819743996e-07,
+      "loss": 0.4277,
+      "step": 11931
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2279753596259868,
+      "learning_rate": 2.883196609095773e-07,
+      "loss": 0.5068,
+      "step": 11932
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1836579056779666,
+      "learning_rate": 2.877209567702011e-07,
+      "loss": 0.5076,
+      "step": 11933
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2130509668289684,
+      "learning_rate": 2.871228658171088e-07,
+      "loss": 0.4813,
+      "step": 11934
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2656736482289208,
+      "learning_rate": 2.8652538808806253e-07,
+      "loss": 0.5475,
+      "step": 11935
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1824683954963262,
+      "learning_rate": 2.8592852362078315e-07,
+      "loss": 0.5113,
+      "step": 11936
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1837851995856066,
+      "learning_rate": 2.85332272452955e-07,
+      "loss": 0.484,
+      "step": 11937
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1781996382082538,
+      "learning_rate": 2.8473663462222025e-07,
+      "loss": 0.5109,
+      "step": 11938
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.161489369386853,
+      "learning_rate": 2.8414161016618757e-07,
+      "loss": 0.474,
+      "step": 11939
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.19865061072465,
+      "learning_rate": 2.8354719912242037e-07,
+      "loss": 0.4294,
+      "step": 11940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2778090349926128,
+      "learning_rate": 2.8295340152845076e-07,
+      "loss": 0.5432,
+      "step": 11941
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2800668510473698,
+      "learning_rate": 2.823602174217643e-07,
+      "loss": 0.5377,
+      "step": 11942
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.340754519166004,
+      "learning_rate": 2.817676468398145e-07,
+      "loss": 0.5583,
+      "step": 11943
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2766522043336932,
+      "learning_rate": 2.811756898200124e-07,
+      "loss": 0.5518,
+      "step": 11944
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1960889002666162,
+      "learning_rate": 2.8058434639973155e-07,
+      "loss": 0.4657,
+      "step": 11945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2276840427821718,
+      "learning_rate": 2.799936166163075e-07,
+      "loss": 0.4676,
+      "step": 11946
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1255634804053662,
+      "learning_rate": 2.79403500507035e-07,
+      "loss": 0.4804,
+      "step": 11947
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1909780264299954,
+      "learning_rate": 2.788139981091698e-07,
+      "loss": 0.4561,
+      "step": 11948
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1905603626557855,
+      "learning_rate": 2.782251094599331e-07,
+      "loss": 0.4857,
+      "step": 11949
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1718303926243183,
+      "learning_rate": 2.7763683459650193e-07,
+      "loss": 0.5005,
+      "step": 11950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2764484872879782,
+      "learning_rate": 2.770491735560199e-07,
+      "loss": 0.4717,
+      "step": 11951
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1982047798765632,
+      "learning_rate": 2.764621263755862e-07,
+      "loss": 0.5143,
+      "step": 11952
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.194503211022297,
+      "learning_rate": 2.758756930922646e-07,
+      "loss": 0.4927,
+      "step": 11953
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1853288578882966,
+      "learning_rate": 2.752898737430809e-07,
+      "loss": 0.5167,
+      "step": 11954
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1788184394264087,
+      "learning_rate": 2.747046683650201e-07,
+      "loss": 0.4516,
+      "step": 11955
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.307880856494758,
+      "learning_rate": 2.741200769950303e-07,
+      "loss": 0.525,
+      "step": 11956
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3012016922836251,
+      "learning_rate": 2.735360996700187e-07,
+      "loss": 0.5012,
+      "step": 11957
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.178059952050663,
+      "learning_rate": 2.7295273642685473e-07,
+      "loss": 0.5073,
+      "step": 11958
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2309126764618947,
+      "learning_rate": 2.7236998730236996e-07,
+      "loss": 0.5447,
+      "step": 11959
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1406729178618724,
+      "learning_rate": 2.71787852333355e-07,
+      "loss": 0.4653,
+      "step": 11960
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1757647744478734,
+      "learning_rate": 2.7120633155656606e-07,
+      "loss": 0.4984,
+      "step": 11961
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.185925845766163,
+      "learning_rate": 2.706254250087126e-07,
+      "loss": 0.4842,
+      "step": 11962
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2629092714518357,
+      "learning_rate": 2.7004513272647415e-07,
+      "loss": 0.4513,
+      "step": 11963
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3132943973643378,
+      "learning_rate": 2.69465454746487e-07,
+      "loss": 0.5086,
+      "step": 11964
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3004609134524585,
+      "learning_rate": 2.688863911053474e-07,
+      "loss": 0.5084,
+      "step": 11965
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2063437667533226,
+      "learning_rate": 2.683079418396173e-07,
+      "loss": 0.5034,
+      "step": 11966
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.208598727097031,
+      "learning_rate": 2.6773010698581516e-07,
+      "loss": 0.5736,
+      "step": 11967
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1611731083505412,
+      "learning_rate": 2.671528865804229e-07,
+      "loss": 0.4801,
+      "step": 11968
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.170193903252669,
+      "learning_rate": 2.6657628065988483e-07,
+      "loss": 0.4723,
+      "step": 11969
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.087732930285538,
+      "learning_rate": 2.6600028926060283e-07,
+      "loss": 0.4972,
+      "step": 11970
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1423176786494544,
+      "learning_rate": 2.6542491241894454e-07,
+      "loss": 0.4859,
+      "step": 11971
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.155597451079196,
+      "learning_rate": 2.648501501712342e-07,
+      "loss": 0.506,
+      "step": 11972
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2153282058183568,
+      "learning_rate": 2.6427600255376164e-07,
+      "loss": 0.4823,
+      "step": 11973
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1563043839320493,
+      "learning_rate": 2.6370246960277344e-07,
+      "loss": 0.4881,
+      "step": 11974
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.167189117798201,
+      "learning_rate": 2.631295513544818e-07,
+      "loss": 0.4691,
+      "step": 11975
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1438665445286416,
+      "learning_rate": 2.625572478450578e-07,
+      "loss": 0.4951,
+      "step": 11976
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3229689650323093,
+      "learning_rate": 2.619855591106324e-07,
+      "loss": 0.546,
+      "step": 11977
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.215251694656694,
+      "learning_rate": 2.614144851873002e-07,
+      "loss": 0.4954,
+      "step": 11978
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.274239550806064,
+      "learning_rate": 2.608440261111178e-07,
+      "loss": 0.5024,
+      "step": 11979
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.273374076153158,
+      "learning_rate": 2.602741819180976e-07,
+      "loss": 0.5202,
+      "step": 11980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1657967673550882,
+      "learning_rate": 2.597049526442197e-07,
+      "loss": 0.5017,
+      "step": 11981
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2373461637342138,
+      "learning_rate": 2.5913633832542083e-07,
+      "loss": 0.5082,
+      "step": 11982
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1556111710310504,
+      "learning_rate": 2.5856833899760123e-07,
+      "loss": 0.4943,
+      "step": 11983
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.176614307818194,
+      "learning_rate": 2.580009546966211e-07,
+      "loss": 0.4251,
+      "step": 11984
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2693501504694302,
+      "learning_rate": 2.574341854583029e-07,
+      "loss": 0.5186,
+      "step": 11985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1877954015226257,
+      "learning_rate": 2.568680313184302e-07,
+      "loss": 0.5017,
+      "step": 11986
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2849889025969692,
+      "learning_rate": 2.563024923127477e-07,
+      "loss": 0.5308,
+      "step": 11987
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.23379657568553,
+      "learning_rate": 2.5573756847695806e-07,
+      "loss": 0.506,
+      "step": 11988
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1914391685065533,
+      "learning_rate": 2.551732598467305e-07,
+      "loss": 0.5038,
+      "step": 11989
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1735511998993036,
+      "learning_rate": 2.5460956645769085e-07,
+      "loss": 0.5073,
+      "step": 11990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0828736809514763,
+      "learning_rate": 2.5404648834542855e-07,
+      "loss": 0.4625,
+      "step": 11991
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2085040837804049,
+      "learning_rate": 2.534840255454962e-07,
+      "loss": 0.5056,
+      "step": 11992
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2284209634914356,
+      "learning_rate": 2.529221780933999e-07,
+      "loss": 0.497,
+      "step": 11993
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.234403972044294,
+      "learning_rate": 2.523609460246168e-07,
+      "loss": 0.518,
+      "step": 11994
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0907697674439658,
+      "learning_rate": 2.5180032937457744e-07,
+      "loss": 0.4483,
+      "step": 11995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2134257685555707,
+      "learning_rate": 2.512403281786768e-07,
+      "loss": 0.4758,
+      "step": 11996
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1679209313443497,
+      "learning_rate": 2.5068094247227227e-07,
+      "loss": 0.4917,
+      "step": 11997
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2417213956987827,
+      "learning_rate": 2.501221722906799e-07,
+      "loss": 0.5062,
+      "step": 11998
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1856591864339103,
+      "learning_rate": 2.4956401766917713e-07,
+      "loss": 0.4469,
+      "step": 11999
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2145081574789132,
+      "learning_rate": 2.4900647864300353e-07,
+      "loss": 0.5042,
+      "step": 12000
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.190122719695915,
+      "learning_rate": 2.4844955524735983e-07,
+      "loss": 0.5048,
+      "step": 12001
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1787707575146524,
+      "learning_rate": 2.4789324751740674e-07,
+      "loss": 0.4698,
+      "step": 12002
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2061319303294689,
+      "learning_rate": 2.4733755548826734e-07,
+      "loss": 0.4878,
+      "step": 12003
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2186741927608036,
+      "learning_rate": 2.467824791950246e-07,
+      "loss": 0.4798,
+      "step": 12004
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2429860745031263,
+      "learning_rate": 2.4622801867272395e-07,
+      "loss": 0.5458,
+      "step": 12005
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.21187530627746,
+      "learning_rate": 2.456741739563717e-07,
+      "loss": 0.4979,
+      "step": 12006
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2818807584502108,
+      "learning_rate": 2.4512094508093553e-07,
+      "loss": 0.512,
+      "step": 12007
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3008086513806378,
+      "learning_rate": 2.445683320813408e-07,
+      "loss": 0.505,
+      "step": 12008
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.165823097883934,
+      "learning_rate": 2.4401633499248065e-07,
+      "loss": 0.5171,
+      "step": 12009
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2662404765517519,
+      "learning_rate": 2.434649538492018e-07,
+      "loss": 0.5399,
+      "step": 12010
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1070861845849567,
+      "learning_rate": 2.4291418868631845e-07,
+      "loss": 0.4199,
+      "step": 12011
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.074086573011623,
+      "learning_rate": 2.423640395386018e-07,
+      "loss": 0.4101,
+      "step": 12012
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3085612278782843,
+      "learning_rate": 2.418145064407862e-07,
+      "loss": 0.5049,
+      "step": 12013
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2489764791553917,
+      "learning_rate": 2.4126558942756617e-07,
+      "loss": 0.5287,
+      "step": 12014
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.248631347960901,
+      "learning_rate": 2.407172885335984e-07,
+      "loss": 0.479,
+      "step": 12015
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.241694466495158,
+      "learning_rate": 2.401696037934997e-07,
+      "loss": 0.5094,
+      "step": 12016
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2000223257054292,
+      "learning_rate": 2.39622535241848e-07,
+      "loss": 0.4584,
+      "step": 12017
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4087392180740177,
+      "learning_rate": 2.3907608291318217e-07,
+      "loss": 0.5408,
+      "step": 12018
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0935510181897552,
+      "learning_rate": 2.3853024684200363e-07,
+      "loss": 0.45,
+      "step": 12019
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2200969686431282,
+      "learning_rate": 2.379850270627726e-07,
+      "loss": 0.5164,
+      "step": 12020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1740251778789763,
+      "learning_rate": 2.3744042360991149e-07,
+      "loss": 0.4775,
+      "step": 12021
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1981639883177975,
+      "learning_rate": 2.3689643651780614e-07,
+      "loss": 0.4422,
+      "step": 12022
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.311206952707571,
+      "learning_rate": 2.363530658207991e-07,
+      "loss": 0.5161,
+      "step": 12023
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3359648049224406,
+      "learning_rate": 2.3581031155319622e-07,
+      "loss": 0.4915,
+      "step": 12024
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2361912945423417,
+      "learning_rate": 2.3526817374926457e-07,
+      "loss": 0.5389,
+      "step": 12025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2609622447104212,
+      "learning_rate": 2.3472665244323346e-07,
+      "loss": 0.5958,
+      "step": 12026
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2986885167861468,
+      "learning_rate": 2.3418574766929215e-07,
+      "loss": 0.4672,
+      "step": 12027
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.194463740175709,
+      "learning_rate": 2.336454594615878e-07,
+      "loss": 0.5082,
+      "step": 12028
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2091560303535505,
+      "learning_rate": 2.331057878542342e-07,
+      "loss": 0.481,
+      "step": 12029
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3155809753010863,
+      "learning_rate": 2.3256673288130194e-07,
+      "loss": 0.5244,
+      "step": 12030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2707554006392356,
+      "learning_rate": 2.3202829457682597e-07,
+      "loss": 0.5044,
+      "step": 12031
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2441987842997217,
+      "learning_rate": 2.314904729748002e-07,
+      "loss": 0.5169,
+      "step": 12032
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3247685374215392,
+      "learning_rate": 2.309532681091786e-07,
+      "loss": 0.5618,
+      "step": 12033
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1598597703228415,
+      "learning_rate": 2.3041668001387852e-07,
+      "loss": 0.4502,
+      "step": 12034
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.146899328571442,
+      "learning_rate": 2.2988070872277834e-07,
+      "loss": 0.4714,
+      "step": 12035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.446801980073105,
+      "learning_rate": 2.293453542697166e-07,
+      "loss": 0.5608,
+      "step": 12036
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3620473092162897,
+      "learning_rate": 2.2881061668849292e-07,
+      "loss": 0.4841,
+      "step": 12037
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2709417401128855,
+      "learning_rate": 2.2827649601286693e-07,
+      "loss": 0.559,
+      "step": 12038
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1602017685922585,
+      "learning_rate": 2.2774299227656282e-07,
+      "loss": 0.4496,
+      "step": 12039
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0776044169235615,
+      "learning_rate": 2.272101055132603e-07,
+      "loss": 0.4796,
+      "step": 12040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2741059029367194,
+      "learning_rate": 2.2667783575660463e-07,
+      "loss": 0.5094,
+      "step": 12041
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2534939096863569,
+      "learning_rate": 2.2614618304020118e-07,
+      "loss": 0.5361,
+      "step": 12042
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1791482160133449,
+      "learning_rate": 2.2561514739761649e-07,
+      "loss": 0.4875,
+      "step": 12043
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2272947033285686,
+      "learning_rate": 2.2508472886237586e-07,
+      "loss": 0.5262,
+      "step": 12044
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1974734052918359,
+      "learning_rate": 2.2455492746796814e-07,
+      "loss": 0.4956,
+      "step": 12045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1009479729454479,
+      "learning_rate": 2.240257432478421e-07,
+      "loss": 0.47,
+      "step": 12046
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2522820146351907,
+      "learning_rate": 2.234971762354099e-07,
+      "loss": 0.5039,
+      "step": 12047
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3355769959294632,
+      "learning_rate": 2.229692264640404e-07,
+      "loss": 0.5158,
+      "step": 12048
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1656456237277526,
+      "learning_rate": 2.2244189396706695e-07,
+      "loss": 0.5079,
+      "step": 12049
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2153635170376567,
+      "learning_rate": 2.219151787777807e-07,
+      "loss": 0.5088,
+      "step": 12050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2055785687228227,
+      "learning_rate": 2.2138908092943834e-07,
+      "loss": 0.5088,
+      "step": 12051
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1917718545243419,
+      "learning_rate": 2.2086360045525444e-07,
+      "loss": 0.4959,
+      "step": 12052
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3766810804509986,
+      "learning_rate": 2.2033873738840584e-07,
+      "loss": 0.4894,
+      "step": 12053
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1321325431876532,
+      "learning_rate": 2.1981449176202818e-07,
+      "loss": 0.4704,
+      "step": 12054
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1921969432131843,
+      "learning_rate": 2.192908636092206e-07,
+      "loss": 0.4869,
+      "step": 12055
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2414569745944832,
+      "learning_rate": 2.1876785296304214e-07,
+      "loss": 0.5467,
+      "step": 12056
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.124624825544048,
+      "learning_rate": 2.1824545985651535e-07,
+      "loss": 0.4943,
+      "step": 12057
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2874571080146386,
+      "learning_rate": 2.1772368432261935e-07,
+      "loss": 0.4755,
+      "step": 12058
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.155986188743347,
+      "learning_rate": 2.1720252639429674e-07,
+      "loss": 0.4736,
+      "step": 12059
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2450519858847429,
+      "learning_rate": 2.1668198610445114e-07,
+      "loss": 0.4918,
+      "step": 12060
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0777794104385683,
+      "learning_rate": 2.1616206348594737e-07,
+      "loss": 0.4503,
+      "step": 12061
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2460140095132886,
+      "learning_rate": 2.1564275857160922e-07,
+      "loss": 0.4804,
+      "step": 12062
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1207008337859716,
+      "learning_rate": 2.151240713942271e-07,
+      "loss": 0.4991,
+      "step": 12063
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2149277930918623,
+      "learning_rate": 2.1460600198654368e-07,
+      "loss": 0.49,
+      "step": 12064
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.17303587317349,
+      "learning_rate": 2.1408855038126953e-07,
+      "loss": 0.4569,
+      "step": 12065
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.21212734709437,
+      "learning_rate": 2.135717166110729e-07,
+      "loss": 0.4933,
+      "step": 12066
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0934133283980239,
+      "learning_rate": 2.1305550070858773e-07,
+      "loss": 0.4728,
+      "step": 12067
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3152276261295004,
+      "learning_rate": 2.1253990270640013e-07,
+      "loss": 0.5402,
+      "step": 12068
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2297246226968586,
+      "learning_rate": 2.1202492263706743e-07,
+      "loss": 0.5255,
+      "step": 12069
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1621308140420006,
+      "learning_rate": 2.1151056053309915e-07,
+      "loss": 0.4744,
+      "step": 12070
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.225138426764048,
+      "learning_rate": 2.1099681642697156e-07,
+      "loss": 0.4856,
+      "step": 12071
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1916113004738296,
+      "learning_rate": 2.104836903511198e-07,
+      "loss": 0.4949,
+      "step": 12072
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3308733608808658,
+      "learning_rate": 2.0997118233794023e-07,
+      "loss": 0.4825,
+      "step": 12073
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1602075225083806,
+      "learning_rate": 2.0945929241978913e-07,
+      "loss": 0.4781,
+      "step": 12074
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.176962832244029,
+      "learning_rate": 2.089480206289851e-07,
+      "loss": 0.5153,
+      "step": 12075
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2696486668207583,
+      "learning_rate": 2.0843736699780792e-07,
+      "loss": 0.4938,
+      "step": 12076
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2157514315170592,
+      "learning_rate": 2.0792733155849842e-07,
+      "loss": 0.5243,
+      "step": 12077
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1521961699558692,
+      "learning_rate": 2.074179143432564e-07,
+      "loss": 0.4574,
+      "step": 12078
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1489113589126132,
+      "learning_rate": 2.0690911538424507e-07,
+      "loss": 0.4659,
+      "step": 12079
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.195690967755979,
+      "learning_rate": 2.0640093471358648e-07,
+      "loss": 0.5091,
+      "step": 12080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.20140458000584,
+      "learning_rate": 2.0589337236336493e-07,
+      "loss": 0.5135,
+      "step": 12081
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3190972869775741,
+      "learning_rate": 2.0538642836562484e-07,
+      "loss": 0.4888,
+      "step": 12082
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2585202234103547,
+      "learning_rate": 2.0488010275237502e-07,
+      "loss": 0.5402,
+      "step": 12083
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3221546391348353,
+      "learning_rate": 2.0437439555557993e-07,
+      "loss": 0.5438,
+      "step": 12084
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.203023782410991,
+      "learning_rate": 2.0386930680716732e-07,
+      "loss": 0.4829,
+      "step": 12085
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2283256153362836,
+      "learning_rate": 2.0336483653902727e-07,
+      "loss": 0.515,
+      "step": 12086
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2224562118738795,
+      "learning_rate": 2.0286098478300986e-07,
+      "loss": 0.4931,
+      "step": 12087
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1949322164137284,
+      "learning_rate": 2.0235775157092407e-07,
+      "loss": 0.4824,
+      "step": 12088
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3721185750380827,
+      "learning_rate": 2.0185513693454338e-07,
+      "loss": 0.5229,
+      "step": 12089
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2677016469903972,
+      "learning_rate": 2.01353140905598e-07,
+      "loss": 0.5052,
+      "step": 12090
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.088415159817075,
+      "learning_rate": 2.0085176351578472e-07,
+      "loss": 0.482,
+      "step": 12091
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2008831390457646,
+      "learning_rate": 2.0035100479675607e-07,
+      "loss": 0.4701,
+      "step": 12092
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3032987135771403,
+      "learning_rate": 1.9985086478012782e-07,
+      "loss": 0.4644,
+      "step": 12093
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2269952095827388,
+      "learning_rate": 1.9935134349747698e-07,
+      "loss": 0.5035,
+      "step": 12094
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2716009552419554,
+      "learning_rate": 1.9885244098034052e-07,
+      "loss": 0.5514,
+      "step": 12095
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1780067748796237,
+      "learning_rate": 1.9835415726021656e-07,
+      "loss": 0.489,
+      "step": 12096
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1294849852652744,
+      "learning_rate": 1.9785649236856442e-07,
+      "loss": 0.4691,
+      "step": 12097
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1899185147155678,
+      "learning_rate": 1.9735944633680448e-07,
+      "loss": 0.4463,
+      "step": 12098
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1943997157886674,
+      "learning_rate": 1.9686301919631833e-07,
+      "loss": 0.4662,
+      "step": 12099
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3422750317915735,
+      "learning_rate": 1.9636721097844648e-07,
+      "loss": 0.573,
+      "step": 12100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1873344004760293,
+      "learning_rate": 1.9587202171449272e-07,
+      "loss": 0.4481,
+      "step": 12101
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2233954684271022,
+      "learning_rate": 1.9537745143572096e-07,
+      "loss": 0.5118,
+      "step": 12102
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3482171259825788,
+      "learning_rate": 1.948835001733551e-07,
+      "loss": 0.4809,
+      "step": 12103
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2038698244123216,
+      "learning_rate": 1.9439016795858357e-07,
+      "loss": 0.5116,
+      "step": 12104
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1548691829894526,
+      "learning_rate": 1.9389745482254918e-07,
+      "loss": 0.5095,
+      "step": 12105
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.187787774001014,
+      "learning_rate": 1.9340536079636263e-07,
+      "loss": 0.5525,
+      "step": 12106
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.199857212551977,
+      "learning_rate": 1.9291388591109017e-07,
+      "loss": 0.5203,
+      "step": 12107
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2188206676633773,
+      "learning_rate": 1.9242303019776253e-07,
+      "loss": 0.5279,
+      "step": 12108
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.220417837756893,
+      "learning_rate": 1.919327936873694e-07,
+      "loss": 0.5136,
+      "step": 12109
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1514269268919335,
+      "learning_rate": 1.9144317641086152e-07,
+      "loss": 0.5104,
+      "step": 12110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3972273977338436,
+      "learning_rate": 1.9095417839915198e-07,
+      "loss": 0.5184,
+      "step": 12111
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2253323843640653,
+      "learning_rate": 1.9046579968311274e-07,
+      "loss": 0.4759,
+      "step": 12112
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2712405860274745,
+      "learning_rate": 1.8997804029357801e-07,
+      "loss": 0.4777,
+      "step": 12113
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2710229179843224,
+      "learning_rate": 1.894909002613432e-07,
+      "loss": 0.4859,
+      "step": 12114
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2055270504157332,
+      "learning_rate": 1.8900437961716257e-07,
+      "loss": 0.5429,
+      "step": 12115
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2961122958722953,
+      "learning_rate": 1.8851847839175375e-07,
+      "loss": 0.5402,
+      "step": 12116
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.28574777457356,
+      "learning_rate": 1.8803319661579554e-07,
+      "loss": 0.515,
+      "step": 12117
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1895949806850161,
+      "learning_rate": 1.8754853431992348e-07,
+      "loss": 0.4661,
+      "step": 12118
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2118984224476805,
+      "learning_rate": 1.870644915347386e-07,
+      "loss": 0.5025,
+      "step": 12119
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.143400447356674,
+      "learning_rate": 1.865810682907987e-07,
+      "loss": 0.4463,
+      "step": 12120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2652369716767213,
+      "learning_rate": 1.8609826461862824e-07,
+      "loss": 0.4942,
+      "step": 12121
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1889098983068962,
+      "learning_rate": 1.8561608054870615e-07,
+      "loss": 0.4918,
+      "step": 12122
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0636012597097029,
+      "learning_rate": 1.8513451611147704e-07,
+      "loss": 0.4285,
+      "step": 12123
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2256293178951103,
+      "learning_rate": 1.8465357133734542e-07,
+      "loss": 0.5635,
+      "step": 12124
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2966667375752512,
+      "learning_rate": 1.8417324625667254e-07,
+      "loss": 0.488,
+      "step": 12125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.220424675273221,
+      "learning_rate": 1.8369354089978643e-07,
+      "loss": 0.4745,
+      "step": 12126
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2795046572076876,
+      "learning_rate": 1.832144552969728e-07,
+      "loss": 0.5058,
+      "step": 12127
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2634688484025987,
+      "learning_rate": 1.827359894784775e-07,
+      "loss": 0.5507,
+      "step": 12128
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1381632223315028,
+      "learning_rate": 1.8225814347451077e-07,
+      "loss": 0.4786,
+      "step": 12129
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1123390359908596,
+      "learning_rate": 1.817809173152396e-07,
+      "loss": 0.4237,
+      "step": 12130
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1959656570186052,
+      "learning_rate": 1.8130431103079437e-07,
+      "loss": 0.505,
+      "step": 12131
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2462354716962607,
+      "learning_rate": 1.8082832465126544e-07,
+      "loss": 0.5069,
+      "step": 12132
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1986920818700946,
+      "learning_rate": 1.803529582067054e-07,
+      "loss": 0.4646,
+      "step": 12133
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3195592322807088,
+      "learning_rate": 1.7987821172712584e-07,
+      "loss": 0.5974,
+      "step": 12134
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.312850950640985,
+      "learning_rate": 1.7940408524249942e-07,
+      "loss": 0.5147,
+      "step": 12135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2052311977810213,
+      "learning_rate": 1.789305787827611e-07,
+      "loss": 0.5016,
+      "step": 12136
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.314242704647338,
+      "learning_rate": 1.7845769237780585e-07,
+      "loss": 0.5303,
+      "step": 12137
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.251032927030242,
+      "learning_rate": 1.779854260574887e-07,
+      "loss": 0.5468,
+      "step": 12138
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2081209467441474,
+      "learning_rate": 1.7751377985162689e-07,
+      "loss": 0.448,
+      "step": 12139
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0903966451479938,
+      "learning_rate": 1.770427537899966e-07,
+      "loss": 0.4337,
+      "step": 12140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2070133738752629,
+      "learning_rate": 1.7657234790233736e-07,
+      "loss": 0.4701,
+      "step": 12141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1493545815011306,
+      "learning_rate": 1.7610256221834765e-07,
+      "loss": 0.506,
+      "step": 12142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2990020406059466,
+      "learning_rate": 1.756333967676882e-07,
+      "loss": 0.5221,
+      "step": 12143
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2151532039675281,
+      "learning_rate": 1.7516485157997975e-07,
+      "loss": 0.5042,
+      "step": 12144
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4123089281328949,
+      "learning_rate": 1.746969266848031e-07,
+      "loss": 0.5706,
+      "step": 12145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2871002058012264,
+      "learning_rate": 1.7422962211170236e-07,
+      "loss": 0.543,
+      "step": 12146
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2048201570809265,
+      "learning_rate": 1.737629378901795e-07,
+      "loss": 0.5779,
+      "step": 12147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1451704304051766,
+      "learning_rate": 1.7329687404969874e-07,
+      "loss": 0.4568,
+      "step": 12148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0856477090044963,
+      "learning_rate": 1.7283143061968654e-07,
+      "loss": 0.4355,
+      "step": 12149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2774099163685264,
+      "learning_rate": 1.7236660762952606e-07,
+      "loss": 0.5469,
+      "step": 12150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2163923886292756,
+      "learning_rate": 1.7190240510856605e-07,
+      "loss": 0.4998,
+      "step": 12151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.180338395783769,
+      "learning_rate": 1.7143882308611305e-07,
+      "loss": 0.4558,
+      "step": 12152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3467759062741584,
+      "learning_rate": 1.7097586159143698e-07,
+      "loss": 0.5154,
+      "step": 12153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2212399210362663,
+      "learning_rate": 1.7051352065376448e-07,
+      "loss": 0.4634,
+      "step": 12154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2382064941175563,
+      "learning_rate": 1.7005180030228886e-07,
+      "loss": 0.4759,
+      "step": 12155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.17365907286405,
+      "learning_rate": 1.695907005661568e-07,
+      "loss": 0.501,
+      "step": 12156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2413998871934906,
+      "learning_rate": 1.691302214744839e-07,
+      "loss": 0.4929,
+      "step": 12157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2592669779072443,
+      "learning_rate": 1.6867036305634022e-07,
+      "loss": 0.5177,
+      "step": 12158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.305444235260479,
+      "learning_rate": 1.6821112534075924e-07,
+      "loss": 0.5249,
+      "step": 12159
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2793911731705891,
+      "learning_rate": 1.6775250835673552e-07,
+      "loss": 0.5262,
+      "step": 12160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.211365458021725,
+      "learning_rate": 1.6729451213322255e-07,
+      "loss": 0.4999,
+      "step": 12161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2238898346042506,
+      "learning_rate": 1.6683713669913836e-07,
+      "loss": 0.5112,
+      "step": 12162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2393212508850167,
+      "learning_rate": 1.6638038208335762e-07,
+      "loss": 0.507,
+      "step": 12163
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2889482852159195,
+      "learning_rate": 1.6592424831471832e-07,
+      "loss": 0.5763,
+      "step": 12164
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.204404819042177,
+      "learning_rate": 1.6546873542201858e-07,
+      "loss": 0.4693,
+      "step": 12165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1430678141510713,
+      "learning_rate": 1.650138434340165e-07,
+      "loss": 0.4736,
+      "step": 12166
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1266679584808017,
+      "learning_rate": 1.6455957237943354e-07,
+      "loss": 0.5101,
+      "step": 12167
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2498853154024294,
+      "learning_rate": 1.6410592228694788e-07,
+      "loss": 0.4867,
+      "step": 12168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1998884586311045,
+      "learning_rate": 1.6365289318520216e-07,
+      "loss": 0.4719,
+      "step": 12169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1959578822511996,
+      "learning_rate": 1.6320048510279906e-07,
+      "loss": 0.4996,
+      "step": 12170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3564610435721631,
+      "learning_rate": 1.6274869806829906e-07,
+      "loss": 0.537,
+      "step": 12171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2130775491906285,
+      "learning_rate": 1.6229753211022825e-07,
+      "loss": 0.4888,
+      "step": 12172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.10593094616037,
+      "learning_rate": 1.6184698725706938e-07,
+      "loss": 0.4739,
+      "step": 12173
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1251039456983958,
+      "learning_rate": 1.613970635372686e-07,
+      "loss": 0.4732,
+      "step": 12174
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.208432812816052,
+      "learning_rate": 1.6094776097923205e-07,
+      "loss": 0.4979,
+      "step": 12175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.228300576120459,
+      "learning_rate": 1.6049907961132595e-07,
+      "loss": 0.5279,
+      "step": 12176
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.178895242183351,
+      "learning_rate": 1.6005101946187873e-07,
+      "loss": 0.4616,
+      "step": 12177
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2592580319743523,
+      "learning_rate": 1.596035805591778e-07,
+      "loss": 0.5268,
+      "step": 12178
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2461039862615906,
+      "learning_rate": 1.5915676293147275e-07,
+      "loss": 0.5007,
+      "step": 12179
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1933419933395568,
+      "learning_rate": 1.5871056660697326e-07,
+      "loss": 0.4544,
+      "step": 12180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2475679102361281,
+      "learning_rate": 1.58264991613849e-07,
+      "loss": 0.5029,
+      "step": 12181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1796509187597046,
+      "learning_rate": 1.5782003798023306e-07,
+      "loss": 0.4684,
+      "step": 12182
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1946990489675904,
+      "learning_rate": 1.5737570573421735e-07,
+      "loss": 0.5202,
+      "step": 12183
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.237524047772037,
+      "learning_rate": 1.56931994903855e-07,
+      "loss": 0.5776,
+      "step": 12184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1674334980487466,
+      "learning_rate": 1.5648890551715924e-07,
+      "loss": 0.4949,
+      "step": 12185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3034596404575625,
+      "learning_rate": 1.5604643760210426e-07,
+      "loss": 0.4857,
+      "step": 12186
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1972559668694358,
+      "learning_rate": 1.5560459118662663e-07,
+      "loss": 0.5006,
+      "step": 12187
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1583195799873276,
+      "learning_rate": 1.5516336629862073e-07,
+      "loss": 0.4864,
+      "step": 12188
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.302249206150199,
+      "learning_rate": 1.5472276296594424e-07,
+      "loss": 0.5478,
+      "step": 12189
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4013798198064138,
+      "learning_rate": 1.5428278121641494e-07,
+      "loss": 0.5301,
+      "step": 12190
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1608281100708564,
+      "learning_rate": 1.538434210778106e-07,
+      "loss": 0.5069,
+      "step": 12191
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3397885414070871,
+      "learning_rate": 1.5340468257787012e-07,
+      "loss": 0.5336,
+      "step": 12192
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2724435121480506,
+      "learning_rate": 1.5296656574429469e-07,
+      "loss": 0.5334,
+      "step": 12193
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.240427365643985,
+      "learning_rate": 1.5252907060474332e-07,
+      "loss": 0.5081,
+      "step": 12194
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1448613234418503,
+      "learning_rate": 1.5209219718683833e-07,
+      "loss": 0.509,
+      "step": 12195
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1004543276525134,
+      "learning_rate": 1.5165594551816209e-07,
+      "loss": 0.4182,
+      "step": 12196
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2538325683241198,
+      "learning_rate": 1.5122031562625593e-07,
+      "loss": 0.5403,
+      "step": 12197
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1693993931748687,
+      "learning_rate": 1.5078530753862453e-07,
+      "loss": 0.4503,
+      "step": 12198
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0690122628360594,
+      "learning_rate": 1.5035092128273144e-07,
+      "loss": 0.4536,
+      "step": 12199
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.287949191130858,
+      "learning_rate": 1.4991715688600362e-07,
+      "loss": 0.5689,
+      "step": 12200
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1418118507533332,
+      "learning_rate": 1.494840143758236e-07,
+      "loss": 0.4651,
+      "step": 12201
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2050147140364351,
+      "learning_rate": 1.4905149377954064e-07,
+      "loss": 0.5358,
+      "step": 12202
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2112102570523997,
+      "learning_rate": 1.4861959512446067e-07,
+      "loss": 0.4725,
+      "step": 12203
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1682220490514175,
+      "learning_rate": 1.4818831843785297e-07,
+      "loss": 0.4725,
+      "step": 12204
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2655820957022974,
+      "learning_rate": 1.4775766374694466e-07,
+      "loss": 0.5276,
+      "step": 12205
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2657145244818975,
+      "learning_rate": 1.4732763107892734e-07,
+      "loss": 0.5174,
+      "step": 12206
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.193764925159402,
+      "learning_rate": 1.4689822046094816e-07,
+      "loss": 0.4801,
+      "step": 12207
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3143625214002959,
+      "learning_rate": 1.4646943192011986e-07,
+      "loss": 0.5459,
+      "step": 12208
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2619091161959535,
+      "learning_rate": 1.4604126548351416e-07,
+      "loss": 0.5323,
+      "step": 12209
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3238760355600567,
+      "learning_rate": 1.4561372117816276e-07,
+      "loss": 0.5621,
+      "step": 12210
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2289468223824649,
+      "learning_rate": 1.451867990310596e-07,
+      "loss": 0.4882,
+      "step": 12211
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2559498328680467,
+      "learning_rate": 1.4476049906915756e-07,
+      "loss": 0.4945,
+      "step": 12212
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1901960386156154,
+      "learning_rate": 1.4433482131937183e-07,
+      "loss": 0.4715,
+      "step": 12213
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2658871508708545,
+      "learning_rate": 1.439097658085764e-07,
+      "loss": 0.5432,
+      "step": 12214
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3043492044220428,
+      "learning_rate": 1.4348533256360985e-07,
+      "loss": 0.4938,
+      "step": 12215
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.274175979766508,
+      "learning_rate": 1.430615216112674e-07,
+      "loss": 0.4766,
+      "step": 12216
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1647408064466906,
+      "learning_rate": 1.426383329783043e-07,
+      "loss": 0.4779,
+      "step": 12217
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2916490902268833,
+      "learning_rate": 1.4221576669144144e-07,
+      "loss": 0.5268,
+      "step": 12218
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3536018024880108,
+      "learning_rate": 1.417938227773563e-07,
+      "loss": 0.5474,
+      "step": 12219
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2916682868865321,
+      "learning_rate": 1.4137250126268876e-07,
+      "loss": 0.5385,
+      "step": 12220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.172899586681365,
+      "learning_rate": 1.409518021740408e-07,
+      "loss": 0.4714,
+      "step": 12221
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2524518761745613,
+      "learning_rate": 1.4053172553797012e-07,
+      "loss": 0.5316,
+      "step": 12222
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2400412581870919,
+      "learning_rate": 1.4011227138099882e-07,
+      "loss": 0.4826,
+      "step": 12223
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.151157712768384,
+      "learning_rate": 1.3969343972961124e-07,
+      "loss": 0.4675,
+      "step": 12224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2691926944908332,
+      "learning_rate": 1.3927523061024962e-07,
+      "loss": 0.538,
+      "step": 12225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2552149232978944,
+      "learning_rate": 1.3885764404931835e-07,
+      "loss": 0.5108,
+      "step": 12226
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1463036467832868,
+      "learning_rate": 1.3844068007317856e-07,
+      "loss": 0.4776,
+      "step": 12227
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2764279877148006,
+      "learning_rate": 1.3802433870815922e-07,
+      "loss": 0.5208,
+      "step": 12228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1551585329751224,
+      "learning_rate": 1.3760861998054264e-07,
+      "loss": 0.4995,
+      "step": 12229
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.244411468967765,
+      "learning_rate": 1.3719352391657893e-07,
+      "loss": 0.5064,
+      "step": 12230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.115744612663472,
+      "learning_rate": 1.3677905054247265e-07,
+      "loss": 0.4775,
+      "step": 12231
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1452109756332838,
+      "learning_rate": 1.363651998843929e-07,
+      "loss": 0.4523,
+      "step": 12232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2152056875493829,
+      "learning_rate": 1.3595197196846655e-07,
+      "loss": 0.4958,
+      "step": 12233
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.38415849770913,
+      "learning_rate": 1.3553936682078494e-07,
+      "loss": 0.5255,
+      "step": 12234
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.298169374841616,
+      "learning_rate": 1.3512738446739726e-07,
+      "loss": 0.5321,
+      "step": 12235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1781939215699344,
+      "learning_rate": 1.347160249343149e-07,
+      "loss": 0.4636,
+      "step": 12236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.242034326862186,
+      "learning_rate": 1.34305288247506e-07,
+      "loss": 0.5068,
+      "step": 12237
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2131952219650455,
+      "learning_rate": 1.3389517443290535e-07,
+      "loss": 0.5173,
+      "step": 12238
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2971394280789845,
+      "learning_rate": 1.3348568351640446e-07,
+      "loss": 0.5145,
+      "step": 12239
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3390528176668712,
+      "learning_rate": 1.3307681552385598e-07,
+      "loss": 0.5027,
+      "step": 12240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1539731160640494,
+      "learning_rate": 1.3266857048107706e-07,
+      "loss": 0.4661,
+      "step": 12241
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0989461901146003,
+      "learning_rate": 1.322609484138382e-07,
+      "loss": 0.4582,
+      "step": 12242
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2298251451866886,
+      "learning_rate": 1.3185394934787766e-07,
+      "loss": 0.5426,
+      "step": 12243
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3323725377826037,
+      "learning_rate": 1.3144757330888934e-07,
+      "loss": 0.5382,
+      "step": 12244
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2823948746068934,
+      "learning_rate": 1.3104182032253164e-07,
+      "loss": 0.5275,
+      "step": 12245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.179933888555367,
+      "learning_rate": 1.3063669041442074e-07,
+      "loss": 0.4926,
+      "step": 12246
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3084780969328043,
+      "learning_rate": 1.302321836101339e-07,
+      "loss": 0.5481,
+      "step": 12247
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2178695139118783,
+      "learning_rate": 1.2982829993521185e-07,
+      "loss": 0.5642,
+      "step": 12248
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1828781331446212,
+      "learning_rate": 1.2942503941515082e-07,
+      "loss": 0.4654,
+      "step": 12249
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2747656326307972,
+      "learning_rate": 1.2902240207541384e-07,
+      "loss": 0.5077,
+      "step": 12250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.194592227714184,
+      "learning_rate": 1.286203879414205e-07,
+      "loss": 0.4848,
+      "step": 12251
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.241201285167893,
+      "learning_rate": 1.2821899703855057e-07,
+      "loss": 0.5149,
+      "step": 12252
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2949184227806783,
+      "learning_rate": 1.2781822939214817e-07,
+      "loss": 0.5862,
+      "step": 12253
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0716363001585496,
+      "learning_rate": 1.2741808502751417e-07,
+      "loss": 0.4715,
+      "step": 12254
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2851832888463306,
+      "learning_rate": 1.2701856396991285e-07,
+      "loss": 0.4644,
+      "step": 12255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3106029060768147,
+      "learning_rate": 1.2661966624456733e-07,
+      "loss": 0.4978,
+      "step": 12256
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.277710188354227,
+      "learning_rate": 1.2622139187666083e-07,
+      "loss": 0.5025,
+      "step": 12257
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1920888477470342,
+      "learning_rate": 1.2582374089134096e-07,
+      "loss": 0.4723,
+      "step": 12258
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1803714209674714,
+      "learning_rate": 1.2542671331371214e-07,
+      "loss": 0.4962,
+      "step": 12259
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1894195000667245,
+      "learning_rate": 1.2503030916884097e-07,
+      "loss": 0.4883,
+      "step": 12260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3001962898689665,
+      "learning_rate": 1.2463452848175516e-07,
+      "loss": 0.5612,
+      "step": 12261
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3101333443283396,
+      "learning_rate": 1.2423937127744146e-07,
+      "loss": 0.5686,
+      "step": 12262
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1750593028422982,
+      "learning_rate": 1.2384483758084765e-07,
+      "loss": 0.4837,
+      "step": 12263
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2637020148761489,
+      "learning_rate": 1.234509274168838e-07,
+      "loss": 0.479,
+      "step": 12264
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1423764303482071,
+      "learning_rate": 1.2305764081042003e-07,
+      "loss": 0.4461,
+      "step": 12265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2452852502232261,
+      "learning_rate": 1.226649777862854e-07,
+      "loss": 0.5491,
+      "step": 12266
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.205153452097077,
+      "learning_rate": 1.2227293836927112e-07,
+      "loss": 0.5095,
+      "step": 12267
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1616423365609658,
+      "learning_rate": 1.2188152258412855e-07,
+      "loss": 0.4445,
+      "step": 12268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.328134424512747,
+      "learning_rate": 1.2149073045557014e-07,
+      "loss": 0.5564,
+      "step": 12269
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2008558400304161,
+      "learning_rate": 1.2110056200826725e-07,
+      "loss": 0.4861,
+      "step": 12270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1278239411164566,
+      "learning_rate": 1.2071101726685464e-07,
+      "loss": 0.4918,
+      "step": 12271
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3167760312658008,
+      "learning_rate": 1.2032209625592705e-07,
+      "loss": 0.5583,
+      "step": 12272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2127233810727611,
+      "learning_rate": 1.199337990000371e-07,
+      "loss": 0.5628,
+      "step": 12273
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3518681704224773,
+      "learning_rate": 1.195461255237007e-07,
+      "loss": 0.5273,
+      "step": 12274
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2611280540741643,
+      "learning_rate": 1.1915907585139385e-07,
+      "loss": 0.4632,
+      "step": 12275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.293948168481518,
+      "learning_rate": 1.1877265000755367e-07,
+      "loss": 0.5143,
+      "step": 12276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1330178962499473,
+      "learning_rate": 1.1838684801657619e-07,
+      "loss": 0.436,
+      "step": 12277
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3154319530940493,
+      "learning_rate": 1.1800166990281858e-07,
+      "loss": 0.5554,
+      "step": 12278
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2347074616251603,
+      "learning_rate": 1.1761711569060141e-07,
+      "loss": 0.4567,
+      "step": 12279
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1541361688837604,
+      "learning_rate": 1.1723318540420081e-07,
+      "loss": 0.4389,
+      "step": 12280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2898600105355889,
+      "learning_rate": 1.1684987906785739e-07,
+      "loss": 0.496,
+      "step": 12281
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2520016379635124,
+      "learning_rate": 1.164671967057729e-07,
+      "loss": 0.5243,
+      "step": 12282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.24934470166198,
+      "learning_rate": 1.1608513834210578e-07,
+      "loss": 0.4977,
+      "step": 12283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2150429320108975,
+      "learning_rate": 1.1570370400097786e-07,
+      "loss": 0.4987,
+      "step": 12284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1670298805693553,
+      "learning_rate": 1.1532289370647209e-07,
+      "loss": 0.4671,
+      "step": 12285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2199556793698574,
+      "learning_rate": 1.1494270748263037e-07,
+      "loss": 0.4974,
+      "step": 12286
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2136574009184797,
+      "learning_rate": 1.1456314535345569e-07,
+      "loss": 0.4839,
+      "step": 12287
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2427437454554302,
+      "learning_rate": 1.1418420734291113e-07,
+      "loss": 0.5363,
+      "step": 12288
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1346942399858833,
+      "learning_rate": 1.1380589347492199e-07,
+      "loss": 0.4042,
+      "step": 12289
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.289938704047202,
+      "learning_rate": 1.134282037733725e-07,
+      "loss": 0.5281,
+      "step": 12290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1204729122307449,
+      "learning_rate": 1.1305113826210911e-07,
+      "loss": 0.4439,
+      "step": 12291
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1737345374148167,
+      "learning_rate": 1.1267469696493726e-07,
+      "loss": 0.4181,
+      "step": 12292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2308029931852427,
+      "learning_rate": 1.1229887990562349e-07,
+      "loss": 0.4944,
+      "step": 12293
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1460344773810023,
+      "learning_rate": 1.1192368710789547e-07,
+      "loss": 0.476,
+      "step": 12294
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1794433341068746,
+      "learning_rate": 1.1154911859543982e-07,
+      "loss": 0.4919,
+      "step": 12295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.121196888289827,
+      "learning_rate": 1.111751743919065e-07,
+      "loss": 0.4683,
+      "step": 12296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1420167766681575,
+      "learning_rate": 1.1080185452090553e-07,
+      "loss": 0.4603,
+      "step": 12297
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.210361959213607,
+      "learning_rate": 1.104291590060036e-07,
+      "loss": 0.4957,
+      "step": 12298
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.239055163059905,
+      "learning_rate": 1.1005708787073189e-07,
+      "loss": 0.53,
+      "step": 12299
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1607990474982268,
+      "learning_rate": 1.096856411385827e-07,
+      "loss": 0.4737,
+      "step": 12300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2387028887862161,
+      "learning_rate": 1.093148188330051e-07,
+      "loss": 0.5061,
+      "step": 12301
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3665108886575255,
+      "learning_rate": 1.0894462097741366e-07,
+      "loss": 0.5371,
+      "step": 12302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2483374984087723,
+      "learning_rate": 1.0857504759517856e-07,
+      "loss": 0.4795,
+      "step": 12303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.141345384503703,
+      "learning_rate": 1.082060987096345e-07,
+      "loss": 0.466,
+      "step": 12304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2948525066408771,
+      "learning_rate": 1.0783777434407394e-07,
+      "loss": 0.5574,
+      "step": 12305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2755014742338422,
+      "learning_rate": 1.0747007452175051e-07,
+      "loss": 0.4813,
+      "step": 12306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.36835800201321,
+      "learning_rate": 1.071029992658823e-07,
+      "loss": 0.5468,
+      "step": 12307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1439187555941306,
+      "learning_rate": 1.0673654859964078e-07,
+      "loss": 0.5287,
+      "step": 12308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1009962642166276,
+      "learning_rate": 1.0637072254616298e-07,
+      "loss": 0.4413,
+      "step": 12309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2153051058406228,
+      "learning_rate": 1.0600552112854712e-07,
+      "loss": 0.513,
+      "step": 12310
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4056457492896397,
+      "learning_rate": 1.0564094436984806e-07,
+      "loss": 0.5283,
+      "step": 12311
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.207458025595689,
+      "learning_rate": 1.0527699229308519e-07,
+      "loss": 0.5225,
+      "step": 12312
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2940130252035187,
+      "learning_rate": 1.0491366492123567e-07,
+      "loss": 0.5125,
+      "step": 12313
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2343347277288745,
+      "learning_rate": 1.0455096227723782e-07,
+      "loss": 0.5343,
+      "step": 12314
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2877700337307048,
+      "learning_rate": 1.0418888438399222e-07,
+      "loss": 0.5395,
+      "step": 12315
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2487096802519029,
+      "learning_rate": 1.0382743126435723e-07,
+      "loss": 0.449,
+      "step": 12316
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1995727811726649,
+      "learning_rate": 1.0346660294115462e-07,
+      "loss": 0.5099,
+      "step": 12317
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.214555859568537,
+      "learning_rate": 1.0310639943716505e-07,
+      "loss": 0.4675,
+      "step": 12318
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2260081988584794,
+      "learning_rate": 1.027468207751292e-07,
+      "loss": 0.4819,
+      "step": 12319
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1420835288510485,
+      "learning_rate": 1.0238786697775006e-07,
+      "loss": 0.4671,
+      "step": 12320
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2172092088127224,
+      "learning_rate": 1.0202953806768945e-07,
+      "loss": 0.4873,
+      "step": 12321
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3107450196155004,
+      "learning_rate": 1.0167183406757042e-07,
+      "loss": 0.518,
+      "step": 12322
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.177888167479233,
+      "learning_rate": 1.0131475499997823e-07,
+      "loss": 0.4877,
+      "step": 12323
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2067505346486307,
+      "learning_rate": 1.0095830088745595e-07,
+      "loss": 0.5093,
+      "step": 12324
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1271417152575818,
+      "learning_rate": 1.0060247175250892e-07,
+      "loss": 0.5121,
+      "step": 12325
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2373623973753018,
+      "learning_rate": 1.0024726761760028e-07,
+      "loss": 0.4876,
+      "step": 12326
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2337505837606269,
+      "learning_rate": 9.989268850515876e-08,
+      "loss": 0.4988,
+      "step": 12327
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1731605599185688,
+      "learning_rate": 9.953873443756979e-08,
+      "loss": 0.4922,
+      "step": 12328
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2897906472118852,
+      "learning_rate": 9.918540543717992e-08,
+      "loss": 0.547,
+      "step": 12329
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.202552907101029,
+      "learning_rate": 9.883270152629686e-08,
+      "loss": 0.5223,
+      "step": 12330
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.218616135899447,
+      "learning_rate": 9.848062272718839e-08,
+      "loss": 0.5378,
+      "step": 12331
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.198193089754735,
+      "learning_rate": 9.812916906208337e-08,
+      "loss": 0.4807,
+      "step": 12332
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2650485197570398,
+      "learning_rate": 9.777834055317181e-08,
+      "loss": 0.5138,
+      "step": 12333
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3114702635755955,
+      "learning_rate": 9.742813722260158e-08,
+      "loss": 0.5745,
+      "step": 12334
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3081470695051902,
+      "learning_rate": 9.707855909248387e-08,
+      "loss": 0.5163,
+      "step": 12335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.217707750740706,
+      "learning_rate": 9.67296061848888e-08,
+      "loss": 0.5089,
+      "step": 12336
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1921186474757552,
+      "learning_rate": 9.638127852184764e-08,
+      "loss": 0.483,
+      "step": 12337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3254214426286584,
+      "learning_rate": 9.60335761253528e-08,
+      "loss": 0.5436,
+      "step": 12338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1557377376413358,
+      "learning_rate": 9.568649901735672e-08,
+      "loss": 0.491,
+      "step": 12339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4133291292277972,
+      "learning_rate": 9.534004721976964e-08,
+      "loss": 0.5685,
+      "step": 12340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3218123815688771,
+      "learning_rate": 9.49942207544674e-08,
+      "loss": 0.5135,
+      "step": 12341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0873486797244585,
+      "learning_rate": 9.464901964328365e-08,
+      "loss": 0.4766,
+      "step": 12342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.391677029888986,
+      "learning_rate": 9.430444390801208e-08,
+      "loss": 0.5173,
+      "step": 12343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2115421385354124,
+      "learning_rate": 9.39604935704086e-08,
+      "loss": 0.4663,
+      "step": 12344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1313327463984348,
+      "learning_rate": 9.361716865218584e-08,
+      "loss": 0.4564,
+      "step": 12345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2953986071414358,
+      "learning_rate": 9.327446917502203e-08,
+      "loss": 0.5349,
+      "step": 12346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2836237991539876,
+      "learning_rate": 9.29323951605532e-08,
+      "loss": 0.4866,
+      "step": 12347
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.236916786093268,
+      "learning_rate": 9.259094663037649e-08,
+      "loss": 0.4894,
+      "step": 12348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1705139903454993,
+      "learning_rate": 9.225012360604802e-08,
+      "loss": 0.4518,
+      "step": 12349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1464493338381698,
+      "learning_rate": 9.190992610908611e-08,
+      "loss": 0.4778,
+      "step": 12350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.187191370962509,
+      "learning_rate": 9.157035416097027e-08,
+      "loss": 0.4557,
+      "step": 12351
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.291838183346705,
+      "learning_rate": 9.123140778313777e-08,
+      "loss": 0.5257,
+      "step": 12352
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.244238354064877,
+      "learning_rate": 9.08930869969904e-08,
+      "loss": 0.4983,
+      "step": 12353
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2552400904182437,
+      "learning_rate": 9.055539182388662e-08,
+      "loss": 0.5225,
+      "step": 12354
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2814056953175004,
+      "learning_rate": 9.021832228514715e-08,
+      "loss": 0.483,
+      "step": 12355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2992610812100918,
+      "learning_rate": 8.988187840205164e-08,
+      "loss": 0.5104,
+      "step": 12356
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1911457715064353,
+      "learning_rate": 8.954606019584312e-08,
+      "loss": 0.5231,
+      "step": 12357
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1331069563593128,
+      "learning_rate": 8.921086768772346e-08,
+      "loss": 0.4375,
+      "step": 12358
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2591847107959995,
+      "learning_rate": 8.887630089885357e-08,
+      "loss": 0.5229,
+      "step": 12359
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2132962787824835,
+      "learning_rate": 8.854235985035875e-08,
+      "loss": 0.4656,
+      "step": 12360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2163328017525223,
+      "learning_rate": 8.820904456331992e-08,
+      "loss": 0.5026,
+      "step": 12361
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2499289492441437,
+      "learning_rate": 8.787635505878245e-08,
+      "loss": 0.536,
+      "step": 12362
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1698546223105282,
+      "learning_rate": 8.754429135775178e-08,
+      "loss": 0.5059,
+      "step": 12363
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.131999862162039,
+      "learning_rate": 8.721285348119113e-08,
+      "loss": 0.4584,
+      "step": 12364
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3023521857855165,
+      "learning_rate": 8.688204145002598e-08,
+      "loss": 0.5216,
+      "step": 12365
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2223730277215143,
+      "learning_rate": 8.655185528514187e-08,
+      "loss": 0.5182,
+      "step": 12366
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.218679181511168,
+      "learning_rate": 8.622229500738655e-08,
+      "loss": 0.4857,
+      "step": 12367
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2070293241318828,
+      "learning_rate": 8.58933606375667e-08,
+      "loss": 0.4279,
+      "step": 12368
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3088704286418436,
+      "learning_rate": 8.556505219644795e-08,
+      "loss": 0.5166,
+      "step": 12369
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3150833682187608,
+      "learning_rate": 8.523736970475927e-08,
+      "loss": 0.4945,
+      "step": 12370
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0519207869756797,
+      "learning_rate": 8.491031318318854e-08,
+      "loss": 0.4668,
+      "step": 12371
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.089009775995349,
+      "learning_rate": 8.458388265238593e-08,
+      "loss": 0.4625,
+      "step": 12372
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2290739843480374,
+      "learning_rate": 8.425807813295939e-08,
+      "loss": 0.5232,
+      "step": 12373
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.071744086720638,
+      "learning_rate": 8.393289964547912e-08,
+      "loss": 0.4929,
+      "step": 12374
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.216458685399548,
+      "learning_rate": 8.360834721047429e-08,
+      "loss": 0.4963,
+      "step": 12375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1171888338094365,
+      "learning_rate": 8.328442084843624e-08,
+      "loss": 0.467,
+      "step": 12376
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.19275806217831,
+      "learning_rate": 8.296112057981643e-08,
+      "loss": 0.5244,
+      "step": 12377
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1741212805663994,
+      "learning_rate": 8.263844642502628e-08,
+      "loss": 0.463,
+      "step": 12378
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2538473525430873,
+      "learning_rate": 8.23163984044395e-08,
+      "loss": 0.4948,
+      "step": 12379
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2948280634205498,
+      "learning_rate": 8.19949765383854e-08,
+      "loss": 0.5345,
+      "step": 12380
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2498698166767315,
+      "learning_rate": 8.167418084715772e-08,
+      "loss": 0.5011,
+      "step": 12381
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.252073189963555,
+      "learning_rate": 8.135401135101251e-08,
+      "loss": 0.4759,
+      "step": 12382
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2972628002141737,
+      "learning_rate": 8.103446807016135e-08,
+      "loss": 0.5339,
+      "step": 12383
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1711803666364815,
+      "learning_rate": 8.071555102478035e-08,
+      "loss": 0.4632,
+      "step": 12384
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1988101225763215,
+      "learning_rate": 8.039726023500227e-08,
+      "loss": 0.5281,
+      "step": 12385
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.18085457374639,
+      "learning_rate": 8.007959572092328e-08,
+      "loss": 0.5131,
+      "step": 12386
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1791846612067254,
+      "learning_rate": 7.976255750260065e-08,
+      "loss": 0.4708,
+      "step": 12387
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.23472831594887,
+      "learning_rate": 7.944614560004838e-08,
+      "loss": 0.5509,
+      "step": 12388
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1848678529670518,
+      "learning_rate": 7.913036003324492e-08,
+      "loss": 0.4916,
+      "step": 12389
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.253816357774344,
+      "learning_rate": 7.881520082212657e-08,
+      "loss": 0.5403,
+      "step": 12390
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2233893783284107,
+      "learning_rate": 7.850066798659072e-08,
+      "loss": 0.5231,
+      "step": 12391
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.231055516408078,
+      "learning_rate": 7.818676154649707e-08,
+      "loss": 0.504,
+      "step": 12392
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3453729277008863,
+      "learning_rate": 7.787348152166197e-08,
+      "loss": 0.5042,
+      "step": 12393
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.194942591656876,
+      "learning_rate": 7.756082793186626e-08,
+      "loss": 0.4935,
+      "step": 12394
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3137209526647342,
+      "learning_rate": 7.724880079684748e-08,
+      "loss": 0.5091,
+      "step": 12395
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.159682514300226,
+      "learning_rate": 7.693740013630768e-08,
+      "loss": 0.5093,
+      "step": 12396
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1842477332510826,
+      "learning_rate": 7.662662596990555e-08,
+      "loss": 0.4457,
+      "step": 12397
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2793293024277577,
+      "learning_rate": 7.631647831726207e-08,
+      "loss": 0.5212,
+      "step": 12398
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0988704170146235,
+      "learning_rate": 7.600695719795936e-08,
+      "loss": 0.4316,
+      "step": 12399
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1589832476901958,
+      "learning_rate": 7.569806263153734e-08,
+      "loss": 0.4934,
+      "step": 12400
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1761044283063211,
+      "learning_rate": 7.53897946375004e-08,
+      "loss": 0.5132,
+      "step": 12401
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2647926984713764,
+      "learning_rate": 7.508215323531076e-08,
+      "loss": 0.4584,
+      "step": 12402
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2613883043418646,
+      "learning_rate": 7.477513844438955e-08,
+      "loss": 0.5084,
+      "step": 12403
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.09665561014378,
+      "learning_rate": 7.446875028412126e-08,
+      "loss": 0.453,
+      "step": 12404
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.114006047470416,
+      "learning_rate": 7.416298877384931e-08,
+      "loss": 0.4832,
+      "step": 12405
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.170540673049112,
+      "learning_rate": 7.385785393287936e-08,
+      "loss": 0.4795,
+      "step": 12406
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2339810092707444,
+      "learning_rate": 7.35533457804749e-08,
+      "loss": 0.5068,
+      "step": 12407
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1914994500399467,
+      "learning_rate": 7.324946433586055e-08,
+      "loss": 0.4664,
+      "step": 12408
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2010054806465922,
+      "learning_rate": 7.29462096182243e-08,
+      "loss": 0.5039,
+      "step": 12409
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2048789281685088,
+      "learning_rate": 7.264358164671082e-08,
+      "loss": 0.5071,
+      "step": 12410
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2081824679848838,
+      "learning_rate": 7.234158044042482e-08,
+      "loss": 0.4874,
+      "step": 12411
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1325617019316214,
+      "learning_rate": 7.204020601843665e-08,
+      "loss": 0.4623,
+      "step": 12412
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2255159517648435,
+      "learning_rate": 7.173945839977103e-08,
+      "loss": 0.4807,
+      "step": 12413
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2583110129458865,
+      "learning_rate": 7.143933760341615e-08,
+      "loss": 0.5206,
+      "step": 12414
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3173090162597205,
+      "learning_rate": 7.113984364832127e-08,
+      "loss": 0.5683,
+      "step": 12415
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2134573529250456,
+      "learning_rate": 7.08409765533935e-08,
+      "loss": 0.4864,
+      "step": 12416
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2282197289660592,
+      "learning_rate": 7.054273633750219e-08,
+      "loss": 0.5317,
+      "step": 12417
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2343168124234116,
+      "learning_rate": 7.024512301947783e-08,
+      "loss": 0.5057,
+      "step": 12418
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2378050071838158,
+      "learning_rate": 6.994813661810984e-08,
+      "loss": 0.4947,
+      "step": 12419
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1777560863106833,
+      "learning_rate": 6.965177715214878e-08,
+      "loss": 0.4611,
+      "step": 12420
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2014387625091465,
+      "learning_rate": 6.93560446403041e-08,
+      "loss": 0.4762,
+      "step": 12421
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1815626386648932,
+      "learning_rate": 6.906093910124756e-08,
+      "loss": 0.4744,
+      "step": 12422
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2697310539885565,
+      "learning_rate": 6.876646055361091e-08,
+      "loss": 0.5249,
+      "step": 12423
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3944536289015632,
+      "learning_rate": 6.847260901598595e-08,
+      "loss": 0.634,
+      "step": 12424
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2221366094939305,
+      "learning_rate": 6.817938450692674e-08,
+      "loss": 0.5356,
+      "step": 12425
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.241579302758814,
+      "learning_rate": 6.788678704494289e-08,
+      "loss": 0.4945,
+      "step": 12426
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.286607520015567,
+      "learning_rate": 6.759481664850853e-08,
+      "loss": 0.5496,
+      "step": 12427
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2600355228835922,
+      "learning_rate": 6.73034733360589e-08,
+      "loss": 0.4979,
+      "step": 12428
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2865756929845544,
+      "learning_rate": 6.70127571259871e-08,
+      "loss": 0.5709,
+      "step": 12429
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.259320273523649,
+      "learning_rate": 6.672266803664729e-08,
+      "loss": 0.4574,
+      "step": 12430
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1934977699358034,
+      "learning_rate": 6.643320608635373e-08,
+      "loss": 0.5093,
+      "step": 12431
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2332053135250418,
+      "learning_rate": 6.614437129338402e-08,
+      "loss": 0.526,
+      "step": 12432
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2581479590756683,
+      "learning_rate": 6.585616367597025e-08,
+      "loss": 0.5149,
+      "step": 12433
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1675335127372417,
+      "learning_rate": 6.556858325231119e-08,
+      "loss": 0.4959,
+      "step": 12434
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2431266161724495,
+      "learning_rate": 6.528163004056231e-08,
+      "loss": 0.5139,
+      "step": 12435
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.265639410996514,
+      "learning_rate": 6.499530405884025e-08,
+      "loss": 0.4859,
+      "step": 12436
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.376424788169291,
+      "learning_rate": 6.470960532522275e-08,
+      "loss": 0.474,
+      "step": 12437
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2535500659294578,
+      "learning_rate": 6.44245338577465e-08,
+      "loss": 0.5081,
+      "step": 12438
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3082010619408375,
+      "learning_rate": 6.414008967441155e-08,
+      "loss": 0.525,
+      "step": 12439
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.204484097617681,
+      "learning_rate": 6.385627279317463e-08,
+      "loss": 0.4792,
+      "step": 12440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.107075685889368,
+      "learning_rate": 6.357308323195476e-08,
+      "loss": 0.4474,
+      "step": 12441
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.349187716075802,
+      "learning_rate": 6.329052100863209e-08,
+      "loss": 0.5518,
+      "step": 12442
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.198955047143343,
+      "learning_rate": 6.300858614104455e-08,
+      "loss": 0.466,
+      "step": 12443
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3348395760546958,
+      "learning_rate": 6.272727864699234e-08,
+      "loss": 0.5237,
+      "step": 12444
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.244967103776308,
+      "learning_rate": 6.244659854423795e-08,
+      "loss": 0.4893,
+      "step": 12445
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1692606437709234,
+      "learning_rate": 6.216654585050052e-08,
+      "loss": 0.4229,
+      "step": 12446
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1444910973285545,
+      "learning_rate": 6.188712058346147e-08,
+      "loss": 0.4876,
+      "step": 12447
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.179135780918476,
+      "learning_rate": 6.160832276076223e-08,
+      "loss": 0.5015,
+      "step": 12448
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1920127950628068,
+      "learning_rate": 6.13301524000054e-08,
+      "loss": 0.4765,
+      "step": 12449
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3099366542555526,
+      "learning_rate": 6.105260951875247e-08,
+      "loss": 0.5431,
+      "step": 12450
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2526664903433207,
+      "learning_rate": 6.077569413452722e-08,
+      "loss": 0.4864,
+      "step": 12451
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.126515744914939,
+      "learning_rate": 6.049940626481121e-08,
+      "loss": 0.4542,
+      "step": 12452
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1824531724944594,
+      "learning_rate": 6.022374592704938e-08,
+      "loss": 0.5002,
+      "step": 12453
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2310202195360875,
+      "learning_rate": 5.994871313864448e-08,
+      "loss": 0.547,
+      "step": 12454
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2322254054532231,
+      "learning_rate": 5.96743079169615e-08,
+      "loss": 0.5195,
+      "step": 12455
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3136923233423534,
+      "learning_rate": 5.9400530279324354e-08,
+      "loss": 0.4679,
+      "step": 12456
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2298715262268494,
+      "learning_rate": 5.9127380243019225e-08,
+      "loss": 0.5101,
+      "step": 12457
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1989055808642282,
+      "learning_rate": 5.88548578252901e-08,
+      "loss": 0.5015,
+      "step": 12458
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2786382246341397,
+      "learning_rate": 5.858296304334321e-08,
+      "loss": 0.474,
+      "step": 12459
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2971925921100096,
+      "learning_rate": 5.831169591434593e-08,
+      "loss": 0.5081,
+      "step": 12460
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.189433932345285,
+      "learning_rate": 5.804105645542235e-08,
+      "loss": 0.4768,
+      "step": 12461
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2423990897143997,
+      "learning_rate": 5.7771044683662125e-08,
+      "loss": 0.5466,
+      "step": 12462
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3103228631986958,
+      "learning_rate": 5.750166061611051e-08,
+      "loss": 0.5119,
+      "step": 12463
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2751552954699292,
+      "learning_rate": 5.7232904269775014e-08,
+      "loss": 0.5203,
+      "step": 12464
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1144996801814857,
+      "learning_rate": 5.696477566162428e-08,
+      "loss": 0.481,
+      "step": 12465
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2273057762987907,
+      "learning_rate": 5.6697274808587e-08,
+      "loss": 0.5185,
+      "step": 12466
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.158732764353542,
+      "learning_rate": 5.6430401727550766e-08,
+      "loss": 0.4804,
+      "step": 12467
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2442779704746576,
+      "learning_rate": 5.6164156435365435e-08,
+      "loss": 0.5379,
+      "step": 12468
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2501807082207745,
+      "learning_rate": 5.5898538948840896e-08,
+      "loss": 0.4778,
+      "step": 12469
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1344399698951528,
+      "learning_rate": 5.563354928474596e-08,
+      "loss": 0.4688,
+      "step": 12470
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.278626570650382,
+      "learning_rate": 5.536918745981168e-08,
+      "loss": 0.5,
+      "step": 12471
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2756372654356487,
+      "learning_rate": 5.5105453490728047e-08,
+      "loss": 0.5128,
+      "step": 12472
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2517831481664743,
+      "learning_rate": 5.484234739414618e-08,
+      "loss": 0.5001,
+      "step": 12473
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1978757215495914,
+      "learning_rate": 5.4579869186676126e-08,
+      "loss": 0.4772,
+      "step": 12474
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1302540267403467,
+      "learning_rate": 5.431801888489241e-08,
+      "loss": 0.515,
+      "step": 12475
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2754454432999216,
+      "learning_rate": 5.405679650532403e-08,
+      "loss": 0.5182,
+      "step": 12476
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.132319218014642,
+      "learning_rate": 5.3796202064464454e-08,
+      "loss": 0.4324,
+      "step": 12477
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.265646004208183,
+      "learning_rate": 5.353623557876608e-08,
+      "loss": 0.4767,
+      "step": 12478
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2585955723394726,
+      "learning_rate": 5.3276897064641344e-08,
+      "loss": 0.5004,
+      "step": 12479
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.165475384721079,
+      "learning_rate": 5.301818653846602e-08,
+      "loss": 0.4736,
+      "step": 12480
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2184190545321012,
+      "learning_rate": 5.27601040165715e-08,
+      "loss": 0.4696,
+      "step": 12481
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1733015913805975,
+      "learning_rate": 5.250264951525364e-08,
+      "loss": 0.5176,
+      "step": 12482
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1832925142492279,
+      "learning_rate": 5.224582305076498e-08,
+      "loss": 0.5312,
+      "step": 12483
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1988818661554093,
+      "learning_rate": 5.198962463932145e-08,
+      "loss": 0.4946,
+      "step": 12484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2209559796714005,
+      "learning_rate": 5.173405429709677e-08,
+      "loss": 0.5064,
+      "step": 12485
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3001314206818908,
+      "learning_rate": 5.147911204022915e-08,
+      "loss": 0.5079,
+      "step": 12486
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2037131618198813,
+      "learning_rate": 5.1224797884812385e-08,
+      "loss": 0.4723,
+      "step": 12487
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1876966915865976,
+      "learning_rate": 5.097111184690251e-08,
+      "loss": 0.5197,
+      "step": 12488
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1044524680634409,
+      "learning_rate": 5.071805394251672e-08,
+      "loss": 0.4765,
+      "step": 12489
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1593487430694185,
+      "learning_rate": 5.046562418763223e-08,
+      "loss": 0.5001,
+      "step": 12490
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3758537936215205,
+      "learning_rate": 5.0213822598185194e-08,
+      "loss": 0.5038,
+      "step": 12491
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1214693616541775,
+      "learning_rate": 4.996264919007399e-08,
+      "loss": 0.4701,
+      "step": 12492
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3261214458782196,
+      "learning_rate": 4.971210397915594e-08,
+      "loss": 0.5007,
+      "step": 12493
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1851359476904144,
+      "learning_rate": 4.9462186981249496e-08,
+      "loss": 0.4835,
+      "step": 12494
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1880002223056367,
+      "learning_rate": 4.9212898212133154e-08,
+      "loss": 0.4855,
+      "step": 12495
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.274575222538688,
+      "learning_rate": 4.8964237687546543e-08,
+      "loss": 0.5616,
+      "step": 12496
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.316144243648743,
+      "learning_rate": 4.871620542318711e-08,
+      "loss": 0.5992,
+      "step": 12497
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2236908748943929,
+      "learning_rate": 4.846880143471677e-08,
+      "loss": 0.5141,
+      "step": 12498
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2364336062205494,
+      "learning_rate": 4.822202573775303e-08,
+      "loss": 0.5309,
+      "step": 12499
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2401378203526143,
+      "learning_rate": 4.797587834787787e-08,
+      "loss": 0.5368,
+      "step": 12500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1809495654856321,
+      "learning_rate": 4.773035928063108e-08,
+      "loss": 0.4937,
+      "step": 12501
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1918283684044366,
+      "learning_rate": 4.748546855151359e-08,
+      "loss": 0.4847,
+      "step": 12502
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.198322420923864,
+      "learning_rate": 4.724120617598637e-08,
+      "loss": 0.4441,
+      "step": 12503
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.226661337982573,
+      "learning_rate": 4.699757216947154e-08,
+      "loss": 0.4981,
+      "step": 12504
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2429011951744944,
+      "learning_rate": 4.675456654735122e-08,
+      "loss": 0.4803,
+      "step": 12505
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1273467700307132,
+      "learning_rate": 4.6512189324966484e-08,
+      "loss": 0.4133,
+      "step": 12506
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3403173560501256,
+      "learning_rate": 4.627044051762064e-08,
+      "loss": 0.5396,
+      "step": 12507
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3218808311532189,
+      "learning_rate": 4.602932014057704e-08,
+      "loss": 0.5339,
+      "step": 12508
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1948304045610205,
+      "learning_rate": 4.5788828209056836e-08,
+      "loss": 0.4882,
+      "step": 12509
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3097140854121707,
+      "learning_rate": 4.5548964738246774e-08,
+      "loss": 0.5099,
+      "step": 12510
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2479411335596724,
+      "learning_rate": 4.530972974328696e-08,
+      "loss": 0.5339,
+      "step": 12511
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2560117163445763,
+      "learning_rate": 4.50711232392842e-08,
+      "loss": 0.5056,
+      "step": 12512
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3385808570359352,
+      "learning_rate": 4.4833145241302e-08,
+      "loss": 0.5483,
+      "step": 12513
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1379624736612282,
+      "learning_rate": 4.4595795764365015e-08,
+      "loss": 0.4174,
+      "step": 12514
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.173764244537531,
+      "learning_rate": 4.4359074823459025e-08,
+      "loss": 0.4725,
+      "step": 12515
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1221474404383212,
+      "learning_rate": 4.412298243352875e-08,
+      "loss": 0.4476,
+      "step": 12516
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2210884157578026,
+      "learning_rate": 4.388751860948004e-08,
+      "loss": 0.5291,
+      "step": 12517
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2707082134418177,
+      "learning_rate": 4.3652683366178784e-08,
+      "loss": 0.4512,
+      "step": 12518
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3398238200048331,
+      "learning_rate": 4.341847671845201e-08,
+      "loss": 0.552,
+      "step": 12519
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3016117869704498,
+      "learning_rate": 4.318489868108677e-08,
+      "loss": 0.5273,
+      "step": 12520
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2658775925239178,
+      "learning_rate": 4.2951949268827955e-08,
+      "loss": 0.5208,
+      "step": 12521
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1328662728010623,
+      "learning_rate": 4.2719628496384894e-08,
+      "loss": 0.4769,
+      "step": 12522
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2399536295649827,
+      "learning_rate": 4.248793637842474e-08,
+      "loss": 0.4867,
+      "step": 12523
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2494636815606133,
+      "learning_rate": 4.22568729295747e-08,
+      "loss": 0.5725,
+      "step": 12524
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1977545034140378,
+      "learning_rate": 4.202643816442309e-08,
+      "loss": 0.5215,
+      "step": 12525
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2078838115135055,
+      "learning_rate": 4.179663209751939e-08,
+      "loss": 0.5288,
+      "step": 12526
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2131186745433011,
+      "learning_rate": 4.156745474337198e-08,
+      "loss": 0.4746,
+      "step": 12527
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2790758713713688,
+      "learning_rate": 4.133890611644931e-08,
+      "loss": 0.5215,
+      "step": 12528
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2232141653862378,
+      "learning_rate": 4.111098623118204e-08,
+      "loss": 0.4987,
+      "step": 12529
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2894984577372375,
+      "learning_rate": 4.0883695101959774e-08,
+      "loss": 0.5045,
+      "step": 12530
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2462082095777405,
+      "learning_rate": 4.0657032743131044e-08,
+      "loss": 0.5209,
+      "step": 12531
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2267800399401552,
+      "learning_rate": 4.0430999169007726e-08,
+      "loss": 0.5071,
+      "step": 12532
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2586559050619879,
+      "learning_rate": 4.0205594393859513e-08,
+      "loss": 0.5277,
+      "step": 12533
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1806697671782416,
+      "learning_rate": 3.9980818431918366e-08,
+      "loss": 0.459,
+      "step": 12534
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2617609353586483,
+      "learning_rate": 3.975667129737515e-08,
+      "loss": 0.5088,
+      "step": 12535
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0412725020744353,
+      "learning_rate": 3.9533153004381873e-08,
+      "loss": 0.4463,
+      "step": 12536
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2060807816003063,
+      "learning_rate": 3.9310263567049476e-08,
+      "loss": 0.5401,
+      "step": 12537
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1969549824655619,
+      "learning_rate": 3.9088002999450034e-08,
+      "loss": 0.5117,
+      "step": 12538
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1970245965041508,
+      "learning_rate": 3.8866371315616766e-08,
+      "loss": 0.4764,
+      "step": 12539
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2294792919982487,
+      "learning_rate": 3.864536852954293e-08,
+      "loss": 0.5033,
+      "step": 12540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1885830307275689,
+      "learning_rate": 3.842499465518068e-08,
+      "loss": 0.5274,
+      "step": 12541
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1805036134450464,
+      "learning_rate": 3.8205249706443345e-08,
+      "loss": 0.4782,
+      "step": 12542
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3933795693284015,
+      "learning_rate": 3.798613369720427e-08,
+      "loss": 0.5868,
+      "step": 12543
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.248727436797367,
+      "learning_rate": 3.776764664129684e-08,
+      "loss": 0.5094,
+      "step": 12544
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1594935621318927,
+      "learning_rate": 3.7549788552517786e-08,
+      "loss": 0.4392,
+      "step": 12545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.411331532355703,
+      "learning_rate": 3.733255944461944e-08,
+      "loss": 0.5604,
+      "step": 12546
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1500280874388213,
+      "learning_rate": 3.711595933131751e-08,
+      "loss": 0.5048,
+      "step": 12547
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.225116241545206,
+      "learning_rate": 3.68999882262866e-08,
+      "loss": 0.4946,
+      "step": 12548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2390220183293383,
+      "learning_rate": 3.668464614316247e-08,
+      "loss": 0.5018,
+      "step": 12549
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3830168406053522,
+      "learning_rate": 3.646993309554092e-08,
+      "loss": 0.508,
+      "step": 12550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.15229367939418,
+      "learning_rate": 3.6255849096976655e-08,
+      "loss": 0.4659,
+      "step": 12551
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2691241271783298,
+      "learning_rate": 3.6042394160987756e-08,
+      "loss": 0.4978,
+      "step": 12552
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2810706966822991,
+      "learning_rate": 3.5829568301049e-08,
+      "loss": 0.473,
+      "step": 12553
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.252853046312049,
+      "learning_rate": 3.561737153059741e-08,
+      "loss": 0.5371,
+      "step": 12554
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3147221781602605,
+      "learning_rate": 3.5405803863032274e-08,
+      "loss": 0.4885,
+      "step": 12555
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1985113685651188,
+      "learning_rate": 3.519486531170735e-08,
+      "loss": 0.5107,
+      "step": 12556
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2007917097203231,
+      "learning_rate": 3.4984555889944204e-08,
+      "loss": 0.4907,
+      "step": 12557
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3001402687545658,
+      "learning_rate": 3.477487561101778e-08,
+      "loss": 0.505,
+      "step": 12558
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2350823994997568,
+      "learning_rate": 3.4565824488166366e-08,
+      "loss": 0.5286,
+      "step": 12559
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.16510285860732,
+      "learning_rate": 3.435740253459052e-08,
+      "loss": 0.5068,
+      "step": 12560
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.14552285004849,
+      "learning_rate": 3.41496097634475e-08,
+      "loss": 0.4835,
+      "step": 12561
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.308576486957611,
+      "learning_rate": 3.3942446187857915e-08,
+      "loss": 0.6124,
+      "step": 12562
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.180858611807922,
+      "learning_rate": 3.373591182089797e-08,
+      "loss": 0.5003,
+      "step": 12563
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2974452867530155,
+      "learning_rate": 3.353000667560946e-08,
+      "loss": 0.4998,
+      "step": 12564
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1624007172386073,
+      "learning_rate": 3.3324730764991985e-08,
+      "loss": 0.4966,
+      "step": 12565
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1742607246251906,
+      "learning_rate": 3.312008410200518e-08,
+      "loss": 0.4783,
+      "step": 12566
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3040481645576383,
+      "learning_rate": 3.2916066699570926e-08,
+      "loss": 0.5205,
+      "step": 12567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.083866092311369,
+      "learning_rate": 3.271267857056781e-08,
+      "loss": 0.5162,
+      "step": 12568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1746480556724765,
+      "learning_rate": 3.250991972783779e-08,
+      "loss": 0.449,
+      "step": 12569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3303790095242969,
+      "learning_rate": 3.230779018418284e-08,
+      "loss": 0.5377,
+      "step": 12570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2407941861421647,
+      "learning_rate": 3.210628995236276e-08,
+      "loss": 0.5397,
+      "step": 12571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0852404765651726,
+      "learning_rate": 3.190541904510069e-08,
+      "loss": 0.4063,
+      "step": 12572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2172157705449325,
+      "learning_rate": 3.170517747507762e-08,
+      "loss": 0.4855,
+      "step": 12573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1867336008575373,
+      "learning_rate": 3.1505565254936755e-08,
+      "loss": 0.5503,
+      "step": 12574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2104968838310863,
+      "learning_rate": 3.130658239728024e-08,
+      "loss": 0.5343,
+      "step": 12575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0946370478706835,
+      "learning_rate": 3.1108228914670245e-08,
+      "loss": 0.4572,
+      "step": 12576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2518267634712072,
+      "learning_rate": 3.0910504819631205e-08,
+      "loss": 0.5355,
+      "step": 12577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2287984986304756,
+      "learning_rate": 3.071341012464535e-08,
+      "loss": 0.5213,
+      "step": 12578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3161426585916085,
+      "learning_rate": 3.051694484215717e-08,
+      "loss": 0.5149,
+      "step": 12579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2385400447508659,
+      "learning_rate": 3.032110898457008e-08,
+      "loss": 0.4832,
+      "step": 12580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2257748644378619,
+      "learning_rate": 3.012590256424752e-08,
+      "loss": 0.5006,
+      "step": 12581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2571520759736823,
+      "learning_rate": 2.993132559351519e-08,
+      "loss": 0.5157,
+      "step": 12582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3422654845349395,
+      "learning_rate": 2.9737378084656597e-08,
+      "loss": 0.4989,
+      "step": 12583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1401826183547275,
+      "learning_rate": 2.954406004991639e-08,
+      "loss": 0.475,
+      "step": 12584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.20654203012724,
+      "learning_rate": 2.935137150150147e-08,
+      "loss": 0.4997,
+      "step": 12585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1993375837364884,
+      "learning_rate": 2.915931245157544e-08,
+      "loss": 0.5235,
+      "step": 12586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.154345051491068,
+      "learning_rate": 2.8967882912265265e-08,
+      "loss": 0.4759,
+      "step": 12587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.212403769459618,
+      "learning_rate": 2.8777082895656837e-08,
+      "loss": 0.4856,
+      "step": 12588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2418752793732681,
+      "learning_rate": 2.8586912413794966e-08,
+      "loss": 0.5041,
+      "step": 12589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3622836859940326,
+      "learning_rate": 2.8397371478687818e-08,
+      "loss": 0.5332,
+      "step": 12590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1319309356028986,
+      "learning_rate": 2.820846010230138e-08,
+      "loss": 0.4785,
+      "step": 12591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2693726895422595,
+      "learning_rate": 2.8020178296562784e-08,
+      "loss": 0.5224,
+      "step": 12592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3295500346937035,
+      "learning_rate": 2.783252607335807e-08,
+      "loss": 0.5083,
+      "step": 12593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1587727321629475,
+      "learning_rate": 2.764550344453554e-08,
+      "loss": 0.507,
+      "step": 12594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1748750052015278,
+      "learning_rate": 2.7459110421903524e-08,
+      "loss": 0.5074,
+      "step": 12595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.128433709424777,
+      "learning_rate": 2.727334701722928e-08,
+      "loss": 0.4599,
+      "step": 12596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2057224325674472,
+      "learning_rate": 2.70882132422412e-08,
+      "loss": 0.4755,
+      "step": 12597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1270810588501117,
+      "learning_rate": 2.6903709108627718e-08,
+      "loss": 0.4236,
+      "step": 12598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3004548634194812,
+      "learning_rate": 2.6719834628037287e-08,
+      "loss": 0.5096,
+      "step": 12599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3358761065023312,
+      "learning_rate": 2.6536589812079517e-08,
+      "loss": 0.5493,
+      "step": 12600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0738701827054769,
+      "learning_rate": 2.6353974672322923e-08,
+      "loss": 0.482,
+      "step": 12601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2054185680097944,
+      "learning_rate": 2.6171989220297177e-08,
+      "loss": 0.4911,
+      "step": 12602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2355684233290045,
+      "learning_rate": 2.5990633467491976e-08,
+      "loss": 0.4885,
+      "step": 12603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.119647272544111,
+      "learning_rate": 2.580990742535705e-08,
+      "loss": 0.4366,
+      "step": 12604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.04430067189396,
+      "learning_rate": 2.562981110530216e-08,
+      "loss": 0.463,
+      "step": 12605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1836326265402604,
+      "learning_rate": 2.545034451869821e-08,
+      "loss": 0.5025,
+      "step": 12606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3094176019188473,
+      "learning_rate": 2.5271507676877248e-08,
+      "loss": 0.5762,
+      "step": 12607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.229317068607476,
+      "learning_rate": 2.5093300591128023e-08,
+      "loss": 0.5105,
+      "step": 12608
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1557713626254547,
+      "learning_rate": 2.4915723272702642e-08,
+      "loss": 0.4225,
+      "step": 12609
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2488192226556767,
+      "learning_rate": 2.4738775732812138e-08,
+      "loss": 0.4733,
+      "step": 12610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.272525765208751,
+      "learning_rate": 2.4562457982628683e-08,
+      "loss": 0.5141,
+      "step": 12611
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.196111075726186,
+      "learning_rate": 2.438677003328338e-08,
+      "loss": 0.5235,
+      "step": 12612
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1794870976886063,
+      "learning_rate": 2.4211711895868462e-08,
+      "loss": 0.4887,
+      "step": 12613
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.168045653806355,
+      "learning_rate": 2.403728358143731e-08,
+      "loss": 0.45,
+      "step": 12614
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1773530206650702,
+      "learning_rate": 2.3863485101001114e-08,
+      "loss": 0.483,
+      "step": 12615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.14599396128836,
+      "learning_rate": 2.3690316465533325e-08,
+      "loss": 0.4959,
+      "step": 12616
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.252387056416989,
+      "learning_rate": 2.3517777685966305e-08,
+      "loss": 0.5047,
+      "step": 12617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2420292879512145,
+      "learning_rate": 2.334586877319467e-08,
+      "loss": 0.5223,
+      "step": 12618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1678464181173436,
+      "learning_rate": 2.3174589738070853e-08,
+      "loss": 0.4807,
+      "step": 12619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2828047088869678,
+      "learning_rate": 2.3003940591408425e-08,
+      "loss": 0.4784,
+      "step": 12620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.144458911679828,
+      "learning_rate": 2.28339213439821e-08,
+      "loss": 0.5012,
+      "step": 12621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2144344903860433,
+      "learning_rate": 2.266453200652552e-08,
+      "loss": 0.4906,
+      "step": 12622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3063654816608978,
+      "learning_rate": 2.2495772589733456e-08,
+      "loss": 0.5064,
+      "step": 12623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2055968122019955,
+      "learning_rate": 2.232764310426072e-08,
+      "loss": 0.5038,
+      "step": 12624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.242079244250955,
+      "learning_rate": 2.2160143560721048e-08,
+      "loss": 0.4767,
+      "step": 12625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3478167009896238,
+      "learning_rate": 2.1993273969691532e-08,
+      "loss": 0.5037,
+      "step": 12626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0968958165293736,
+      "learning_rate": 2.1827034341704855e-08,
+      "loss": 0.4541,
+      "step": 12627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2237881427154937,
+      "learning_rate": 2.166142468725929e-08,
+      "loss": 0.5013,
+      "step": 12628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1549132068440089,
+      "learning_rate": 2.1496445016809808e-08,
+      "loss": 0.4535,
+      "step": 12629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1663659877688526,
+      "learning_rate": 2.133209534077141e-08,
+      "loss": 0.4821,
+      "step": 12630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3071388013406051,
+      "learning_rate": 2.1168375669521346e-08,
+      "loss": 0.5582,
+      "step": 12631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2073892599468885,
+      "learning_rate": 2.1005286013394688e-08,
+      "loss": 0.4629,
+      "step": 12632
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2083539906193934,
+      "learning_rate": 2.0842826382689864e-08,
+      "loss": 0.4557,
+      "step": 12633
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.154444806064512,
+      "learning_rate": 2.068099678766311e-08,
+      "loss": 0.4534,
+      "step": 12634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2691488306232885,
+      "learning_rate": 2.0519797238531813e-08,
+      "loss": 0.5131,
+      "step": 12635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1546102701874512,
+      "learning_rate": 2.0359227745472278e-08,
+      "loss": 0.4635,
+      "step": 12636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.22282886200742,
+      "learning_rate": 2.0199288318624165e-08,
+      "loss": 0.5263,
+      "step": 12637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2637600757412284,
+      "learning_rate": 2.003997896808274e-08,
+      "loss": 0.513,
+      "step": 12638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3187990260841724,
+      "learning_rate": 1.988129970390773e-08,
+      "loss": 0.4691,
+      "step": 12639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.220886363200298,
+      "learning_rate": 1.972325053611779e-08,
+      "loss": 0.4486,
+      "step": 12640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1272970696167466,
+      "learning_rate": 1.9565831474689377e-08,
+      "loss": 0.5075,
+      "step": 12641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2195276078262047,
+      "learning_rate": 1.9409042529562327e-08,
+      "loss": 0.4991,
+      "step": 12642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.165288548324413,
+      "learning_rate": 1.9252883710635383e-08,
+      "loss": 0.5002,
+      "step": 12643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.257544163104745,
+      "learning_rate": 1.909735502776844e-08,
+      "loss": 0.563,
+      "step": 12644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1696889196413247,
+      "learning_rate": 1.8942456490780305e-08,
+      "loss": 0.4523,
+      "step": 12645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3571724861538996,
+      "learning_rate": 1.8788188109449822e-08,
+      "loss": 0.5102,
+      "step": 12646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.232620247474563,
+      "learning_rate": 1.8634549893516983e-08,
+      "loss": 0.4757,
+      "step": 12647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3154257453570524,
+      "learning_rate": 1.8481541852682917e-08,
+      "loss": 0.5532,
+      "step": 12648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1616027240495295,
+      "learning_rate": 1.832916399660656e-08,
+      "loss": 0.4633,
+      "step": 12649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.192816877909245,
+      "learning_rate": 1.8177416334907995e-08,
+      "loss": 0.5134,
+      "step": 12650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1525956218050522,
+      "learning_rate": 1.8026298877169557e-08,
+      "loss": 0.4664,
+      "step": 12651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1901307329124517,
+      "learning_rate": 1.7875811632930285e-08,
+      "loss": 0.4796,
+      "step": 12652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1985796489738958,
+      "learning_rate": 1.7725954611692577e-08,
+      "loss": 0.4977,
+      "step": 12653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2222830108254599,
+      "learning_rate": 1.757672782291775e-08,
+      "loss": 0.4819,
+      "step": 12654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.142257097579135,
+      "learning_rate": 1.742813127602605e-08,
+      "loss": 0.5035,
+      "step": 12655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.11314546526645,
+      "learning_rate": 1.728016498039886e-08,
+      "loss": 0.4671,
+      "step": 12656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2919311509019973,
+      "learning_rate": 1.713282894537982e-08,
+      "loss": 0.5232,
+      "step": 12657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1539355646850773,
+      "learning_rate": 1.6986123180270376e-08,
+      "loss": 0.4443,
+      "step": 12658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2444507924884904,
+      "learning_rate": 1.6840047694332007e-08,
+      "loss": 0.4721,
+      "step": 12659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2600276704050837,
+      "learning_rate": 1.6694602496788447e-08,
+      "loss": 0.481,
+      "step": 12660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1196517975268374,
+      "learning_rate": 1.6549787596821242e-08,
+      "loss": 0.4641,
+      "step": 12661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2072081698534942,
+      "learning_rate": 1.640560300357308e-08,
+      "loss": 0.5326,
+      "step": 12662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2229784946094853,
+      "learning_rate": 1.6262048726148894e-08,
+      "loss": 0.5128,
+      "step": 12663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.297595134400607,
+      "learning_rate": 1.6119124773610328e-08,
+      "loss": 0.4879,
+      "step": 12664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3088338147772178,
+      "learning_rate": 1.5976831154981275e-08,
+      "loss": 0.4948,
+      "step": 12665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2027569988287352,
+      "learning_rate": 1.583516787924566e-08,
+      "loss": 0.4992,
+      "step": 12666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.211887946450703,
+      "learning_rate": 1.569413495534744e-08,
+      "loss": 0.5026,
+      "step": 12667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.212820299846413,
+      "learning_rate": 1.5553732392191712e-08,
+      "loss": 0.5573,
+      "step": 12668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1885641750619365,
+      "learning_rate": 1.5413960198641388e-08,
+      "loss": 0.511,
+      "step": 12669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.283391140383113,
+      "learning_rate": 1.527481838352052e-08,
+      "loss": 0.525,
+      "step": 12670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1577687470152433,
+      "learning_rate": 1.513630695561541e-08,
+      "loss": 0.5019,
+      "step": 12671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1802811802432251,
+      "learning_rate": 1.499842592367018e-08,
+      "loss": 0.4751,
+      "step": 12672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.184738763592111,
+      "learning_rate": 1.4861175296390084e-08,
+      "loss": 0.4256,
+      "step": 12673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1066242554783943,
+      "learning_rate": 1.4724555082441528e-08,
+      "loss": 0.4851,
+      "step": 12674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2692062666224042,
+      "learning_rate": 1.458856529044872e-08,
+      "loss": 0.5065,
+      "step": 12675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.233754738563783,
+      "learning_rate": 1.4453205928997016e-08,
+      "loss": 0.5212,
+      "step": 12676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2134549460580195,
+      "learning_rate": 1.4318477006632914e-08,
+      "loss": 0.5174,
+      "step": 12677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1839502886601159,
+      "learning_rate": 1.418437853186294e-08,
+      "loss": 0.488,
+      "step": 12678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.243352283933351,
+      "learning_rate": 1.405091051315366e-08,
+      "loss": 0.4615,
+      "step": 12679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2655166297009866,
+      "learning_rate": 1.3918072958931662e-08,
+      "loss": 0.4869,
+      "step": 12680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2254379851869046,
+      "learning_rate": 1.3785865877581351e-08,
+      "loss": 0.4764,
+      "step": 12681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2155500609533092,
+      "learning_rate": 1.3654289277452715e-08,
+      "loss": 0.4928,
+      "step": 12682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2407765562561746,
+      "learning_rate": 1.3523343166851332e-08,
+      "loss": 0.5097,
+      "step": 12683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2287690548665986,
+      "learning_rate": 1.3393027554045035e-08,
+      "loss": 0.4988,
+      "step": 12684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2577501364465047,
+      "learning_rate": 1.3263342447260575e-08,
+      "loss": 0.455,
+      "step": 12685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.304038154599678,
+      "learning_rate": 1.3134287854685846e-08,
+      "loss": 0.4923,
+      "step": 12686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.106439602223515,
+      "learning_rate": 1.3005863784468774e-08,
+      "loss": 0.4866,
+      "step": 12687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2633800611854116,
+      "learning_rate": 1.2878070244718433e-08,
+      "loss": 0.5417,
+      "step": 12688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.186966776340299,
+      "learning_rate": 1.2750907243501698e-08,
+      "loss": 0.4708,
+      "step": 12689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2494317193488416,
+      "learning_rate": 1.2624374788848814e-08,
+      "loss": 0.4961,
+      "step": 12690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3156602596532554,
+      "learning_rate": 1.2498472888745616e-08,
+      "loss": 0.4772,
+      "step": 12691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2698992386592534,
+      "learning_rate": 1.2373201551143521e-08,
+      "loss": 0.5202,
+      "step": 12692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1804663001084925,
+      "learning_rate": 1.2248560783950646e-08,
+      "loss": 0.5111,
+      "step": 12693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2526284239938883,
+      "learning_rate": 1.2124550595036255e-08,
+      "loss": 0.5367,
+      "step": 12694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2306195363494494,
+      "learning_rate": 1.2001170992228528e-08,
+      "loss": 0.4877,
+      "step": 12695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.177407746072349,
+      "learning_rate": 1.187842198331901e-08,
+      "loss": 0.5063,
+      "step": 12696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.181903248756463,
+      "learning_rate": 1.175630357605706e-08,
+      "loss": 0.4957,
+      "step": 12697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.109665523688508,
+      "learning_rate": 1.1634815778150954e-08,
+      "loss": 0.5012,
+      "step": 12698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1468266198870443,
+      "learning_rate": 1.1513958597273445e-08,
+      "loss": 0.4636,
+      "step": 12699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3306020637423956,
+      "learning_rate": 1.1393732041052874e-08,
+      "loss": 0.5574,
+      "step": 12700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.231717593095607,
+      "learning_rate": 1.1274136117080946e-08,
+      "loss": 0.501,
+      "step": 12701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2513176172452023,
+      "learning_rate": 1.115517083290718e-08,
+      "loss": 0.5012,
+      "step": 12702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1676995723145023,
+      "learning_rate": 1.1036836196043344e-08,
+      "loss": 0.496,
+      "step": 12703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2389142557482795,
+      "learning_rate": 1.0919132213960126e-08,
+      "loss": 0.5206,
+      "step": 12704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1331752327895728,
+      "learning_rate": 1.0802058894089363e-08,
+      "loss": 0.4722,
+      "step": 12705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1956660925767462,
+      "learning_rate": 1.0685616243821806e-08,
+      "loss": 0.4979,
+      "step": 12706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2624269747623686,
+      "learning_rate": 1.0569804270509354e-08,
+      "loss": 0.5209,
+      "step": 12707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2372739527617718,
+      "learning_rate": 1.0454622981463935e-08,
+      "loss": 0.5085,
+      "step": 12708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4619405638306442,
+      "learning_rate": 1.034007238395751e-08,
+      "loss": 0.5298,
+      "step": 12709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2525100778414118,
+      "learning_rate": 1.0226152485222073e-08,
+      "loss": 0.4629,
+      "step": 12710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2862428286457241,
+      "learning_rate": 1.0112863292450758e-08,
+      "loss": 0.5394,
+      "step": 12711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0701279864016118,
+      "learning_rate": 1.0000204812794511e-08,
+      "loss": 0.4706,
+      "step": 12712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.163185608610436,
+      "learning_rate": 9.888177053367642e-09,
+      "loss": 0.4616,
+      "step": 12713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2559422395934243,
+      "learning_rate": 9.776780021241161e-09,
+      "loss": 0.524,
+      "step": 12714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1807718409210963,
+      "learning_rate": 9.666013723450552e-09,
+      "loss": 0.4642,
+      "step": 12715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.314801106296906,
+      "learning_rate": 9.555878166987998e-09,
+      "loss": 0.525,
+      "step": 12716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2418551210253237,
+      "learning_rate": 9.446373358805716e-09,
+      "loss": 0.5123,
+      "step": 12717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2614900836471672,
+      "learning_rate": 9.337499305819287e-09,
+      "loss": 0.499,
+      "step": 12718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4273683676348017,
+      "learning_rate": 9.229256014900989e-09,
+      "loss": 0.5802,
+      "step": 12719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1766008327769735,
+      "learning_rate": 9.121643492885358e-09,
+      "loss": 0.4721,
+      "step": 12720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1745903599693428,
+      "learning_rate": 9.014661746566954e-09,
+      "loss": 0.4524,
+      "step": 12721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.348019139179454,
+      "learning_rate": 8.90831078269927e-09,
+      "loss": 0.5121,
+      "step": 12722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.301335809390326,
+      "learning_rate": 8.80259060799804e-09,
+      "loss": 0.5051,
+      "step": 12723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1327991353266051,
+      "learning_rate": 8.697501229135708e-09,
+      "loss": 0.4515,
+      "step": 12724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1477310706795962,
+      "learning_rate": 8.593042652749184e-09,
+      "loss": 0.4624,
+      "step": 12725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3856188213122878,
+      "learning_rate": 8.489214885433195e-09,
+      "loss": 0.5233,
+      "step": 12726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2141359972216665,
+      "learning_rate": 8.386017933741386e-09,
+      "loss": 0.4808,
+      "step": 12727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2309942183122504,
+      "learning_rate": 8.283451804190767e-09,
+      "loss": 0.5559,
+      "step": 12728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1809277614687603,
+      "learning_rate": 8.18151650325616e-09,
+      "loss": 0.4801,
+      "step": 12729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3661456252677353,
+      "learning_rate": 8.080212037374636e-09,
+      "loss": 0.54,
+      "step": 12730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2866423111094476,
+      "learning_rate": 7.97953841293997e-09,
+      "loss": 0.5094,
+      "step": 12731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2295635951048012,
+      "learning_rate": 7.879495636308188e-09,
+      "loss": 0.5078,
+      "step": 12732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2228714141184835,
+      "learning_rate": 7.78008371379757e-09,
+      "loss": 0.4787,
+      "step": 12733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2432130622646405,
+      "learning_rate": 7.681302651683098e-09,
+      "loss": 0.5176,
+      "step": 12734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2042672815934001,
+      "learning_rate": 7.583152456200892e-09,
+      "loss": 0.4908,
+      "step": 12735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2644028590890768,
+      "learning_rate": 7.485633133549331e-09,
+      "loss": 0.5024,
+      "step": 12736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2938829860431587,
+      "learning_rate": 7.3887446898834865e-09,
+      "loss": 0.4993,
+      "step": 12737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3347526784106538,
+      "learning_rate": 7.292487131321802e-09,
+      "loss": 0.5021,
+      "step": 12738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1829435873019603,
+      "learning_rate": 7.1968604639416354e-09,
+      "loss": 0.4752,
+      "step": 12739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2377566119691414,
+      "learning_rate": 7.1018646937781596e-09,
+      "loss": 0.4571,
+      "step": 12740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.228612316796155,
+      "learning_rate": 7.007499826832132e-09,
+      "loss": 0.4611,
+      "step": 12741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2366804009743393,
+      "learning_rate": 6.913765869058786e-09,
+      "loss": 0.4754,
+      "step": 12742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.182124973370936,
+      "learning_rate": 6.820662826376723e-09,
+      "loss": 0.5172,
+      "step": 12743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2616431624030153,
+      "learning_rate": 6.728190704664572e-09,
+      "loss": 0.4868,
+      "step": 12744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.230429754698932,
+      "learning_rate": 6.636349509760997e-09,
+      "loss": 0.5769,
+      "step": 12745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1692607457236275,
+      "learning_rate": 6.545139247462473e-09,
+      "loss": 0.4374,
+      "step": 12746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1358010694175988,
+      "learning_rate": 6.454559923529946e-09,
+      "loss": 0.4582,
+      "step": 12747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2873372873245224,
+      "learning_rate": 6.364611543679955e-09,
+      "loss": 0.5132,
+      "step": 12748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.331729172268932,
+      "learning_rate": 6.275294113592401e-09,
+      "loss": 0.535,
+      "step": 12749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1422743695123125,
+      "learning_rate": 6.186607638907216e-09,
+      "loss": 0.4593,
+      "step": 12750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2191683222476324,
+      "learning_rate": 6.098552125222146e-09,
+      "loss": 0.516,
+      "step": 12751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1540944394717993,
+      "learning_rate": 6.0111275780971865e-09,
+      "loss": 0.4492,
+      "step": 12752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2025172692722335,
+      "learning_rate": 5.924334003052368e-09,
+      "loss": 0.5226,
+      "step": 12753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2681666625780705,
+      "learning_rate": 5.838171405566639e-09,
+      "loss": 0.5018,
+      "step": 12754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0935665521729219,
+      "learning_rate": 5.752639791080095e-09,
+      "loss": 0.4704,
+      "step": 12755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1377924930832464,
+      "learning_rate": 5.667739164993968e-09,
+      "loss": 0.4579,
+      "step": 12756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3449772398203494,
+      "learning_rate": 5.583469532666197e-09,
+      "loss": 0.4832,
+      "step": 12757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1875041660436534,
+      "learning_rate": 5.49983089941919e-09,
+      "loss": 0.4417,
+      "step": 12758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1745423034253408,
+      "learning_rate": 5.416823270532057e-09,
+      "loss": 0.495,
+      "step": 12759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2282987806532002,
+      "learning_rate": 5.334446651246161e-09,
+      "loss": 0.487,
+      "step": 12760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.342449889997135,
+      "learning_rate": 5.252701046762898e-09,
+      "loss": 0.5254,
+      "step": 12761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2715543619281244,
+      "learning_rate": 5.171586462242584e-09,
+      "loss": 0.4865,
+      "step": 12762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.094478745872354,
+      "learning_rate": 5.091102902806677e-09,
+      "loss": 0.4544,
+      "step": 12763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3019438808787112,
+      "learning_rate": 5.011250373535559e-09,
+      "loss": 0.4927,
+      "step": 12764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.298966892258081,
+      "learning_rate": 4.932028879472972e-09,
+      "loss": 0.5164,
+      "step": 12765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1492545406906582,
+      "learning_rate": 4.8534384256182506e-09,
+      "loss": 0.4609,
+      "step": 12766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1557608936018389,
+      "learning_rate": 4.775479016934092e-09,
+      "loss": 0.5113,
+      "step": 12767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0462360424901878,
+      "learning_rate": 4.698150658343226e-09,
+      "loss": 0.4544,
+      "step": 12768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2507010877989286,
+      "learning_rate": 4.621453354726191e-09,
+      "loss": 0.5325,
+      "step": 12769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2383028142838413,
+      "learning_rate": 4.545387110926891e-09,
+      "loss": 0.4787,
+      "step": 12770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1392576827492746,
+      "learning_rate": 4.4699519317459305e-09,
+      "loss": 0.4477,
+      "step": 12771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2822288402760853,
+      "learning_rate": 4.395147821948387e-09,
+      "loss": 0.5515,
+      "step": 12772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2124480147868137,
+      "learning_rate": 4.320974786254928e-09,
+      "loss": 0.4641,
+      "step": 12773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2341956962057914,
+      "learning_rate": 4.247432829349585e-09,
+      "loss": 0.5244,
+      "step": 12774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2746961025762316,
+      "learning_rate": 4.17452195587531e-09,
+      "loss": 0.5264,
+      "step": 12775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1431327842918724,
+      "learning_rate": 4.102242170435089e-09,
+      "loss": 0.4609,
+      "step": 12776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2277499240105445,
+      "learning_rate": 4.030593477591938e-09,
+      "loss": 0.4699,
+      "step": 12777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.207457087684253,
+      "learning_rate": 3.959575881870015e-09,
+      "loss": 0.4785,
+      "step": 12778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2098685684230215,
+      "learning_rate": 3.889189387752401e-09,
+      "loss": 0.477,
+      "step": 12779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1490477413322115,
+      "learning_rate": 3.81943399968443e-09,
+      "loss": 0.4662,
+      "step": 12780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.15566826697599,
+      "learning_rate": 3.750309722068135e-09,
+      "loss": 0.4696,
+      "step": 12781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1612448674346205,
+      "learning_rate": 3.6818165592689138e-09,
+      "loss": 0.4977,
+      "step": 12782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.32667039987304,
+      "learning_rate": 3.6139545156110845e-09,
+      "loss": 0.5464,
+      "step": 12783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1690986296927226,
+      "learning_rate": 3.546723595378998e-09,
+      "loss": 0.5062,
+      "step": 12784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.242103237916961,
+      "learning_rate": 3.480123802817037e-09,
+      "loss": 0.4797,
+      "step": 12785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1727203371959878,
+      "learning_rate": 3.4141551421296158e-09,
+      "loss": 0.5447,
+      "step": 12786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2458725975168847,
+      "learning_rate": 3.348817617483402e-09,
+      "loss": 0.5028,
+      "step": 12787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3535245644630611,
+      "learning_rate": 3.284111233000653e-09,
+      "loss": 0.5261,
+      "step": 12788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2896032412979295,
+      "learning_rate": 3.2200359927692105e-09,
+      "loss": 0.476,
+      "step": 12789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1653521261845836,
+      "learning_rate": 3.1565919008336164e-09,
+      "loss": 0.5147,
+      "step": 12790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1364300023553127,
+      "learning_rate": 3.093778961199556e-09,
+      "loss": 0.4568,
+      "step": 12791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.212805506964488,
+      "learning_rate": 3.0315971778316354e-09,
+      "loss": 0.4889,
+      "step": 12792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2480915760040547,
+      "learning_rate": 2.9700465546567136e-09,
+      "loss": 0.5756,
+      "step": 12793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2684883881497169,
+      "learning_rate": 2.909127095560571e-09,
+      "loss": 0.5234,
+      "step": 12794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1934599638905135,
+      "learning_rate": 2.8488388043901303e-09,
+      "loss": 0.475,
+      "step": 12795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2552156355818238,
+      "learning_rate": 2.7891816849501264e-09,
+      "loss": 0.4996,
+      "step": 12796
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.365215077625375,
+      "learning_rate": 2.7301557410086554e-09,
+      "loss": 0.5418,
+      "step": 12797
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1764671884866285,
+      "learning_rate": 2.671760976291626e-09,
+      "loss": 0.4622,
+      "step": 12798
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2295036285086227,
+      "learning_rate": 2.6139973944849795e-09,
+      "loss": 0.469,
+      "step": 12799
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1373493765571152,
+      "learning_rate": 2.556864999236908e-09,
+      "loss": 0.5109,
+      "step": 12800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1620364382639425,
+      "learning_rate": 2.5003637941534153e-09,
+      "loss": 0.4847,
+      "step": 12801
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2873899764669479,
+      "learning_rate": 2.4444937828027592e-09,
+      "loss": 0.5283,
+      "step": 12802
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.285990018455841,
+      "learning_rate": 2.389254968712118e-09,
+      "loss": 0.4971,
+      "step": 12803
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.141489511282108,
+      "learning_rate": 2.334647355368702e-09,
+      "loss": 0.4104,
+      "step": 12804
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2558244430406191,
+      "learning_rate": 2.280670946219754e-09,
+      "loss": 0.4653,
+      "step": 12805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2653039595715418,
+      "learning_rate": 2.2273257446736586e-09,
+      "loss": 0.4555,
+      "step": 12806
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.212344625683869,
+      "learning_rate": 2.174611754097722e-09,
+      "loss": 0.5237,
+      "step": 12807
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.325821033750812,
+      "learning_rate": 2.122528977821503e-09,
+      "loss": 0.5268,
+      "step": 12808
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2797011340932254,
+      "learning_rate": 2.071077419131262e-09,
+      "loss": 0.5498,
+      "step": 12809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.177253941462722,
+      "learning_rate": 2.0202570812777323e-09,
+      "loss": 0.4783,
+      "step": 12810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.234470846583571,
+      "learning_rate": 1.9700679674672373e-09,
+      "loss": 0.5146,
+      "step": 12811
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2242221737286103,
+      "learning_rate": 1.9205100808694644e-09,
+      "loss": 0.5137,
+      "step": 12812
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.185292249607657,
+      "learning_rate": 1.8715834246130215e-09,
+      "loss": 0.5357,
+      "step": 12813
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2970198584854682,
+      "learning_rate": 1.8232880017876597e-09,
+      "loss": 0.5064,
+      "step": 12814
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.283950936137628,
+      "learning_rate": 1.775623815440941e-09,
+      "loss": 0.5173,
+      "step": 12815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1472506456238225,
+      "learning_rate": 1.7285908685837904e-09,
+      "loss": 0.4178,
+      "step": 12816
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1838934490613877,
+      "learning_rate": 1.6821891641860543e-09,
+      "loss": 0.4775,
+      "step": 12817
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3355287517736054,
+      "learning_rate": 1.636418705174281e-09,
+      "loss": 0.5806,
+      "step": 12818
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2837279943197735,
+      "learning_rate": 1.591279494441711e-09,
+      "loss": 0.5243,
+      "step": 12819
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.258090871015429,
+      "learning_rate": 1.5467715348360668e-09,
+      "loss": 0.4976,
+      "step": 12820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1104370191663984,
+      "learning_rate": 1.502894829167323e-09,
+      "loss": 0.4364,
+      "step": 12821
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2667553410527597,
+      "learning_rate": 1.459649380207706e-09,
+      "loss": 0.4724,
+      "step": 12822
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.313096274307284,
+      "learning_rate": 1.4170351906850344e-09,
+      "loss": 0.537,
+      "step": 12823
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1776148290993629,
+      "learning_rate": 1.3750522632915986e-09,
+      "loss": 0.468,
+      "step": 12824
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4211429765678127,
+      "learning_rate": 1.3337006006763908e-09,
+      "loss": 0.5084,
+      "step": 12825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0702948464378694,
+      "learning_rate": 1.292980205451766e-09,
+      "loss": 0.4506,
+      "step": 12826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3529927618590143,
+      "learning_rate": 1.25289108018678e-09,
+      "loss": 0.5111,
+      "step": 12827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3467819695085206,
+      "learning_rate": 1.2134332274149619e-09,
+      "loss": 0.5082,
+      "step": 12828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1463209617389374,
+      "learning_rate": 1.1746066496243214e-09,
+      "loss": 0.4773,
+      "step": 12829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2291410034000652,
+      "learning_rate": 1.1364113492695617e-09,
+      "loss": 0.4777,
+      "step": 12830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1750131930970078,
+      "learning_rate": 1.0988473287598667e-09,
+      "loss": 0.521,
+      "step": 12831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3122985321913427,
+      "learning_rate": 1.061914590467783e-09,
+      "loss": 0.5624,
+      "step": 12832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.193020936515233,
+      "learning_rate": 1.0256131367236688e-09,
+      "loss": 0.5322,
+      "step": 12833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2552957412022692,
+      "learning_rate": 9.899429698212448e-10,
+      "loss": 0.4719,
+      "step": 12834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1478150945312688,
+      "learning_rate": 9.549040920120433e-10,
+      "loss": 0.4432,
+      "step": 12835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1489883451210987,
+      "learning_rate": 9.204965055076287e-10,
+      "loss": 0.49,
+      "step": 12836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2036015941369709,
+      "learning_rate": 8.867202124818175e-10,
+      "loss": 0.4963,
+      "step": 12837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1304358445277378,
+      "learning_rate": 8.535752150651277e-10,
+      "loss": 0.4901,
+      "step": 12838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2300702130816425,
+      "learning_rate": 8.210615153503299e-10,
+      "loss": 0.5489,
+      "step": 12839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1910347781544575,
+      "learning_rate": 7.891791153924466e-10,
+      "loss": 0.5424,
+      "step": 12840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1025158159155233,
+      "learning_rate": 7.579280172020919e-10,
+      "loss": 0.4601,
+      "step": 12841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.13989950648964,
+      "learning_rate": 7.273082227532424e-10,
+      "loss": 0.5087,
+      "step": 12842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2908781731593217,
+      "learning_rate": 6.97319733977686e-10,
+      "loss": 0.5117,
+      "step": 12843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1354390180061806,
+      "learning_rate": 6.679625527716837e-10,
+      "loss": 0.4893,
+      "step": 12844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3138743425755401,
+      "learning_rate": 6.392366809859773e-10,
+      "loss": 0.5462,
+      "step": 12845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2061990381257566,
+      "learning_rate": 6.111421204357814e-10,
+      "loss": 0.4884,
+      "step": 12846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2542096302788721,
+      "learning_rate": 5.836788728930121e-10,
+      "loss": 0.5311,
+      "step": 12847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3202656393369088,
+      "learning_rate": 5.568469400940579e-10,
+      "loss": 0.5222,
+      "step": 12848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3227679652146,
+      "learning_rate": 5.30646323730899e-10,
+      "loss": 0.5325,
+      "step": 12849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2876945867063403,
+      "learning_rate": 5.050770254588777e-10,
+      "loss": 0.5751,
+      "step": 12850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.187574032935247,
+      "learning_rate": 4.801390468922584e-10,
+      "loss": 0.4579,
+      "step": 12851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.285203463258182,
+      "learning_rate": 4.5583238960533736e-10,
+      "loss": 0.5233,
+      "step": 12852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1133538473451257,
+      "learning_rate": 4.321570551324428e-10,
+      "loss": 0.4226,
+      "step": 12853
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1663889838318557,
+      "learning_rate": 4.091130449679348e-10,
+      "loss": 0.4677,
+      "step": 12854
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1724266788193352,
+      "learning_rate": 3.8670036056731585e-10,
+      "loss": 0.493,
+      "step": 12855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.13859785568944,
+      "learning_rate": 3.649190033461203e-10,
+      "loss": 0.4663,
+      "step": 12856
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2017358463883105,
+      "learning_rate": 3.43768974677694e-10,
+      "loss": 0.4671,
+      "step": 12857
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3005809916172948,
+      "learning_rate": 3.232502758998557e-10,
+      "loss": 0.5257,
+      "step": 12858
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1712508001497808,
+      "learning_rate": 3.0336290830601523e-10,
+      "loss": 0.4861,
+      "step": 12859
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1428841466353052,
+      "learning_rate": 2.8410687315294504e-10,
+      "loss": 0.4977,
+      "step": 12860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3101388037287138,
+      "learning_rate": 2.6548217165633937e-10,
+      "loss": 0.5179,
+      "step": 12861
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1900309144340138,
+      "learning_rate": 2.474888049908142e-10,
+      "loss": 0.4883,
+      "step": 12862
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0761382552916974,
+      "learning_rate": 2.3012677429323782e-10,
+      "loss": 0.4621,
+      "step": 12863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1498276996075498,
+      "learning_rate": 2.1339608066051064e-10,
+      "loss": 0.5184,
+      "step": 12864
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2162790436526083,
+      "learning_rate": 1.9729672514845477e-10,
+      "loss": 0.4903,
+      "step": 12865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.364325050002346,
+      "learning_rate": 1.81828708771814e-10,
+      "loss": 0.5696,
+      "step": 12866
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.210383036020965,
+      "learning_rate": 1.669920325098051e-10,
+      "loss": 0.5131,
+      "step": 12867
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2000195938735925,
+      "learning_rate": 1.5278669729723582e-10,
+      "loss": 0.4936,
+      "step": 12868
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.17990206352433,
+      "learning_rate": 1.392127040322766e-10,
+      "loss": 0.497,
+      "step": 12869
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1404167403217738,
+      "learning_rate": 1.2627005357090938e-10,
+      "loss": 0.4577,
+      "step": 12870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2433683912194202,
+      "learning_rate": 1.1395874673136853e-10,
+      "loss": 0.5056,
+      "step": 12871
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2172539160338953,
+      "learning_rate": 1.0227878428969995e-10,
+      "loss": 0.4555,
+      "step": 12872
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2806800411938544,
+      "learning_rate": 9.123016698309173e-11,
+      "loss": 0.4979,
+      "step": 12873
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2302234223442006,
+      "learning_rate": 8.081289551098437e-11,
+      "loss": 0.4998,
+      "step": 12874
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.265595094303915,
+      "learning_rate": 7.102697052951968e-11,
+      "loss": 0.4699,
+      "step": 12875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3867426050847587,
+      "learning_rate": 6.187239265709188e-11,
+      "loss": 0.5241,
+      "step": 12876
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.126374358869901,
+      "learning_rate": 5.334916247212718e-11,
+      "loss": 0.4993,
+      "step": 12877
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2527189248455288,
+      "learning_rate": 4.5457280510863286e-11,
+      "loss": 0.5114,
+      "step": 12878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2873742810691888,
+      "learning_rate": 3.8196747274010794e-11,
+      "loss": 0.5341,
+      "step": 12879
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.073346480907056,
+      "learning_rate": 3.1567563217871356e-11,
+      "loss": 0.4682,
+      "step": 12880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.216398121752131,
+      "learning_rate": 2.5569728763219505e-11,
+      "loss": 0.4824,
+      "step": 12881
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.237128506395995,
+      "learning_rate": 2.0203244286420842e-11,
+      "loss": 0.498,
+      "step": 12882
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2252380606881614,
+      "learning_rate": 1.546811012720362e-11,
+      "loss": 0.5131,
+      "step": 12883
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1420296681286557,
+      "learning_rate": 1.1364326585328045e-11,
+      "loss": 0.4729,
+      "step": 12884
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1551170468989012,
+      "learning_rate": 7.891893918365868e-12,
+      "loss": 0.4921,
+      "step": 12885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5112252625112261,
+      "learning_rate": 5.0508123472514655e-12,
+      "loss": 0.4912,
+      "step": 12886
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1867831726870262,
+      "learning_rate": 2.841082049620525e-12,
+      "loss": 0.5362,
+      "step": 12887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3153736807745786,
+      "learning_rate": 1.2627031664713685e-12,
+      "loss": 0.5467,
+      "step": 12888
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3907085779341495,
+      "learning_rate": 3.156757966138457e-13,
+      "loss": 0.5349,
+      "step": 12889
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3798019914655004,
+      "learning_rate": 0.0,
+      "loss": 0.4959,
+      "step": 12890
+    },
+    {
+      "epoch": 1.0,
+      "step": 12890,
+      "total_flos": 4.01818824385495e+16,
+      "train_loss": 0.5671361568664931,
+      "train_runtime": 271490.8709,
+      "train_samples_per_second": 6.077,
+      "train_steps_per_second": 0.047
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12890,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 700,
+  "total_flos": 4.01818824385495e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}