diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4430 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.80384,
+  "eval_steps": 500,
+  "global_step": 628,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00128,
+      "grad_norm": 2.8870697066158506,
+      "learning_rate": 0.0,
+      "loss": 0.8422,
+      "step": 1
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 2.88484389829891,
+      "learning_rate": 6.329113924050633e-07,
+      "loss": 0.8541,
+      "step": 2
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 2.858151965789657,
+      "learning_rate": 1.2658227848101265e-06,
+      "loss": 0.8376,
+      "step": 3
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 2.759628117182127,
+      "learning_rate": 1.8987341772151901e-06,
+      "loss": 0.8334,
+      "step": 4
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 2.796990062811218,
+      "learning_rate": 2.531645569620253e-06,
+      "loss": 0.8256,
+      "step": 5
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 2.5779298795445023,
+      "learning_rate": 3.1645569620253167e-06,
+      "loss": 0.8301,
+      "step": 6
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 2.182261607936066,
+      "learning_rate": 3.7974683544303802e-06,
+      "loss": 0.8156,
+      "step": 7
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 1.9615896152651355,
+      "learning_rate": 4.430379746835443e-06,
+      "loss": 0.7982,
+      "step": 8
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 1.452541644948315,
+      "learning_rate": 5.063291139240506e-06,
+      "loss": 0.7819,
+      "step": 9
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 1.4723286808630864,
+      "learning_rate": 5.69620253164557e-06,
+      "loss": 0.7906,
+      "step": 10
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 1.3529636617858944,
+      "learning_rate": 6.329113924050633e-06,
+      "loss": 0.7724,
+      "step": 11
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 1.960737179905222,
+      "learning_rate": 6.9620253164556965e-06,
+      "loss": 0.7495,
+      "step": 12
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 2.2349101055406337,
+      "learning_rate": 7.5949367088607605e-06,
+      "loss": 0.7581,
+      "step": 13
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 2.0897577150322477,
+      "learning_rate": 8.227848101265822e-06,
+      "loss": 0.7404,
+      "step": 14
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 1.8227218322635887,
+      "learning_rate": 8.860759493670886e-06,
+      "loss": 0.7382,
+      "step": 15
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 1.2099951464458898,
+      "learning_rate": 9.49367088607595e-06,
+      "loss": 0.7231,
+      "step": 16
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 1.2177037129914572,
+      "learning_rate": 1.0126582278481012e-05,
+      "loss": 0.7259,
+      "step": 17
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 1.1031346132830708,
+      "learning_rate": 1.0759493670886076e-05,
+      "loss": 0.7059,
+      "step": 18
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 0.9194779600801882,
+      "learning_rate": 1.139240506329114e-05,
+      "loss": 0.7137,
+      "step": 19
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.8679468005972053,
+      "learning_rate": 1.2025316455696203e-05,
+      "loss": 0.7036,
+      "step": 20
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 0.7227287276969042,
+      "learning_rate": 1.2658227848101267e-05,
+      "loss": 0.696,
+      "step": 21
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 0.7425882516811844,
+      "learning_rate": 1.3291139240506329e-05,
+      "loss": 0.6888,
+      "step": 22
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 0.7093793012252196,
+      "learning_rate": 1.3924050632911393e-05,
+      "loss": 0.6791,
+      "step": 23
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 0.6018215463147907,
+      "learning_rate": 1.4556962025316457e-05,
+      "loss": 0.6783,
+      "step": 24
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.5846346732378257,
+      "learning_rate": 1.5189873417721521e-05,
+      "loss": 0.6811,
+      "step": 25
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 0.5855419788452784,
+      "learning_rate": 1.5822784810126583e-05,
+      "loss": 0.683,
+      "step": 26
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 0.5096689891724868,
+      "learning_rate": 1.6455696202531644e-05,
+      "loss": 0.6589,
+      "step": 27
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 0.4871170504081146,
+      "learning_rate": 1.7088607594936708e-05,
+      "loss": 0.6582,
+      "step": 28
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 0.4949600697144217,
+      "learning_rate": 1.7721518987341772e-05,
+      "loss": 0.669,
+      "step": 29
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 0.5082926031630941,
+      "learning_rate": 1.8354430379746836e-05,
+      "loss": 0.666,
+      "step": 30
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 0.49381475380567175,
+      "learning_rate": 1.89873417721519e-05,
+      "loss": 0.6556,
+      "step": 31
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 0.4265624784331274,
+      "learning_rate": 1.962025316455696e-05,
+      "loss": 0.646,
+      "step": 32
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 0.39190416547723717,
+      "learning_rate": 2.0253164556962025e-05,
+      "loss": 0.6473,
+      "step": 33
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 0.4631353399929371,
+      "learning_rate": 2.088607594936709e-05,
+      "loss": 0.6441,
+      "step": 34
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 0.3928335126997034,
+      "learning_rate": 2.1518987341772153e-05,
+      "loss": 0.6352,
+      "step": 35
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 0.36295027582313966,
+      "learning_rate": 2.2151898734177217e-05,
+      "loss": 0.6333,
+      "step": 36
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 0.35026852064181846,
+      "learning_rate": 2.278481012658228e-05,
+      "loss": 0.6399,
+      "step": 37
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 0.39778614916835536,
+      "learning_rate": 2.341772151898734e-05,
+      "loss": 0.6298,
+      "step": 38
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 0.33278348666417684,
+      "learning_rate": 2.4050632911392405e-05,
+      "loss": 0.6301,
+      "step": 39
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 0.31444068712551376,
+      "learning_rate": 2.468354430379747e-05,
+      "loss": 0.6263,
+      "step": 40
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 0.36059728676958264,
+      "learning_rate": 2.5316455696202533e-05,
+      "loss": 0.6458,
+      "step": 41
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 0.3916144552301749,
+      "learning_rate": 2.5949367088607597e-05,
+      "loss": 0.6331,
+      "step": 42
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 0.32338566356420756,
+      "learning_rate": 2.6582278481012658e-05,
+      "loss": 0.6332,
+      "step": 43
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 0.33704233729853356,
+      "learning_rate": 2.7215189873417722e-05,
+      "loss": 0.6348,
+      "step": 44
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.36015399213900634,
+      "learning_rate": 2.7848101265822786e-05,
+      "loss": 0.6392,
+      "step": 45
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 0.31471331803021757,
+      "learning_rate": 2.848101265822785e-05,
+      "loss": 0.6272,
+      "step": 46
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 0.3225170654156012,
+      "learning_rate": 2.9113924050632914e-05,
+      "loss": 0.635,
+      "step": 47
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 0.3064473735810606,
+      "learning_rate": 2.9746835443037974e-05,
+      "loss": 0.6284,
+      "step": 48
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 0.3038289969291092,
+      "learning_rate": 3.0379746835443042e-05,
+      "loss": 0.6149,
+      "step": 49
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.3226803690164346,
+      "learning_rate": 3.10126582278481e-05,
+      "loss": 0.626,
+      "step": 50
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 0.3096398144524693,
+      "learning_rate": 3.1645569620253167e-05,
+      "loss": 0.621,
+      "step": 51
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 0.2754757429130796,
+      "learning_rate": 3.227848101265823e-05,
+      "loss": 0.6185,
+      "step": 52
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 0.3262507218160328,
+      "learning_rate": 3.291139240506329e-05,
+      "loss": 0.6171,
+      "step": 53
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 0.34971068352090656,
+      "learning_rate": 3.354430379746836e-05,
+      "loss": 0.616,
+      "step": 54
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 0.2841621281043231,
+      "learning_rate": 3.4177215189873416e-05,
+      "loss": 0.5995,
+      "step": 55
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 0.4003223636484448,
+      "learning_rate": 3.4810126582278487e-05,
+      "loss": 0.6169,
+      "step": 56
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 0.31868860231705426,
+      "learning_rate": 3.5443037974683544e-05,
+      "loss": 0.6077,
+      "step": 57
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 0.3960425782005289,
+      "learning_rate": 3.607594936708861e-05,
+      "loss": 0.6164,
+      "step": 58
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 0.363865574596696,
+      "learning_rate": 3.670886075949367e-05,
+      "loss": 0.6118,
+      "step": 59
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.33961478774466697,
+      "learning_rate": 3.7341772151898736e-05,
+      "loss": 0.6137,
+      "step": 60
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 0.4212164741206082,
+      "learning_rate": 3.79746835443038e-05,
+      "loss": 0.6275,
+      "step": 61
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 0.29878729710395663,
+      "learning_rate": 3.8607594936708864e-05,
+      "loss": 0.6084,
+      "step": 62
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 0.36745026817379894,
+      "learning_rate": 3.924050632911392e-05,
+      "loss": 0.607,
+      "step": 63
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 0.38983571508393644,
+      "learning_rate": 3.987341772151899e-05,
+      "loss": 0.6176,
+      "step": 64
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.37337392917475115,
+      "learning_rate": 4.050632911392405e-05,
+      "loss": 0.6184,
+      "step": 65
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 0.3668068115925863,
+      "learning_rate": 4.113924050632912e-05,
+      "loss": 0.6194,
+      "step": 66
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 0.36138503055306903,
+      "learning_rate": 4.177215189873418e-05,
+      "loss": 0.6077,
+      "step": 67
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 0.43361127462043814,
+      "learning_rate": 4.240506329113924e-05,
+      "loss": 0.6147,
+      "step": 68
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 0.33520423726109644,
+      "learning_rate": 4.3037974683544305e-05,
+      "loss": 0.6118,
+      "step": 69
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.4381154362148859,
+      "learning_rate": 4.367088607594937e-05,
+      "loss": 0.6031,
+      "step": 70
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 0.3717345864324632,
+      "learning_rate": 4.430379746835443e-05,
+      "loss": 0.6031,
+      "step": 71
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 0.4861728465398392,
+      "learning_rate": 4.49367088607595e-05,
+      "loss": 0.6006,
+      "step": 72
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 0.3264992939190504,
+      "learning_rate": 4.556962025316456e-05,
+      "loss": 0.6151,
+      "step": 73
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 0.4319794925001871,
+      "learning_rate": 4.6202531645569625e-05,
+      "loss": 0.6058,
+      "step": 74
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.4616345840492333,
+      "learning_rate": 4.683544303797468e-05,
+      "loss": 0.5967,
+      "step": 75
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 0.4405721152587957,
+      "learning_rate": 4.7468354430379746e-05,
+      "loss": 0.6002,
+      "step": 76
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 0.5122605377853799,
+      "learning_rate": 4.810126582278481e-05,
+      "loss": 0.6076,
+      "step": 77
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 0.45313870340097556,
+      "learning_rate": 4.8734177215189874e-05,
+      "loss": 0.6074,
+      "step": 78
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 0.4340044755876676,
+      "learning_rate": 4.936708860759494e-05,
+      "loss": 0.606,
+      "step": 79
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.4987172862476422,
+      "learning_rate": 5e-05,
+      "loss": 0.6158,
+      "step": 80
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 0.6226880208665108,
+      "learning_rate": 4.999974965737065e-05,
+      "loss": 0.621,
+      "step": 81
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 0.5448293131914782,
+      "learning_rate": 4.999899863449631e-05,
+      "loss": 0.6014,
+      "step": 82
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 0.3427022601926917,
+      "learning_rate": 4.999774694641803e-05,
+      "loss": 0.6198,
+      "step": 83
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 0.5005152113593655,
+      "learning_rate": 4.999599461820387e-05,
+      "loss": 0.6054,
+      "step": 84
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.5702968806820528,
+      "learning_rate": 4.999374168494844e-05,
+      "loss": 0.6069,
+      "step": 85
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 0.4671310661706222,
+      "learning_rate": 4.999098819177214e-05,
+      "loss": 0.6017,
+      "step": 86
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 0.46081768174689064,
+      "learning_rate": 4.9987734193820324e-05,
+      "loss": 0.5988,
+      "step": 87
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 0.5448729856183013,
+      "learning_rate": 4.9983979756262136e-05,
+      "loss": 0.6181,
+      "step": 88
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 0.5095775592779056,
+      "learning_rate": 4.9979724954289244e-05,
+      "loss": 0.608,
+      "step": 89
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.41119162739543413,
+      "learning_rate": 4.997496987311431e-05,
+      "loss": 0.5979,
+      "step": 90
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 0.45501958535738946,
+      "learning_rate": 4.996971460796929e-05,
+      "loss": 0.6019,
+      "step": 91
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 0.4287172104360816,
+      "learning_rate": 4.9963959264103544e-05,
+      "loss": 0.5955,
+      "step": 92
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.409872269342458,
+      "learning_rate": 4.995770395678171e-05,
+      "loss": 0.5927,
+      "step": 93
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 0.4304173966206036,
+      "learning_rate": 4.995094881128138e-05,
+      "loss": 0.5967,
+      "step": 94
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.4229799776298517,
+      "learning_rate": 4.994369396289063e-05,
+      "loss": 0.6084,
+      "step": 95
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 0.4509596954971553,
+      "learning_rate": 4.9935939556905295e-05,
+      "loss": 0.6134,
+      "step": 96
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 0.467661146414229,
+      "learning_rate": 4.992768574862603e-05,
+      "loss": 0.5986,
+      "step": 97
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 0.42432875998240194,
+      "learning_rate": 4.9918932703355256e-05,
+      "loss": 0.6028,
+      "step": 98
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.43479377184835605,
+      "learning_rate": 4.990968059639379e-05,
+      "loss": 0.5942,
+      "step": 99
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.3680676685801686,
+      "learning_rate": 4.989992961303738e-05,
+      "loss": 0.5994,
+      "step": 100
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 0.3956815409903461,
+      "learning_rate": 4.9889679948572974e-05,
+      "loss": 0.5871,
+      "step": 101
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.34354949934586104,
+      "learning_rate": 4.98789318082748e-05,
+      "loss": 0.5873,
+      "step": 102
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 0.3608260963951222,
+      "learning_rate": 4.986768540740028e-05,
+      "loss": 0.5883,
+      "step": 103
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 0.3937004101078116,
+      "learning_rate": 4.98559409711857e-05,
+      "loss": 0.6029,
+      "step": 104
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.3401718481532899,
+      "learning_rate": 4.9843698734841705e-05,
+      "loss": 0.5983,
+      "step": 105
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 0.4371868869288284,
+      "learning_rate": 4.983095894354858e-05,
+      "loss": 0.5866,
+      "step": 106
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 0.3722813571279646,
+      "learning_rate": 4.981772185245135e-05,
+      "loss": 0.5954,
+      "step": 107
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.36493596395606354,
+      "learning_rate": 4.980398772665468e-05,
+      "loss": 0.5806,
+      "step": 108
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 0.43678937522389644,
+      "learning_rate": 4.9789756841217546e-05,
+      "loss": 0.595,
+      "step": 109
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.34968596729530604,
+      "learning_rate": 4.977502948114772e-05,
+      "loss": 0.5999,
+      "step": 110
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 0.4035249077012057,
+      "learning_rate": 4.9759805941396075e-05,
+      "loss": 0.582,
+      "step": 111
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 0.3396387531525401,
+      "learning_rate": 4.974408652685072e-05,
+      "loss": 0.5912,
+      "step": 112
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 0.3888124435581031,
+      "learning_rate": 4.9727871552330794e-05,
+      "loss": 0.5994,
+      "step": 113
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.3487289265208422,
+      "learning_rate": 4.971116134258025e-05,
+      "loss": 0.598,
+      "step": 114
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.34084258932596606,
+      "learning_rate": 4.969395623226133e-05,
+      "loss": 0.5965,
+      "step": 115
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 0.33211872605390524,
+      "learning_rate": 4.967625656594782e-05,
+      "loss": 0.5984,
+      "step": 116
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.31055192632357626,
+      "learning_rate": 4.9658062698118213e-05,
+      "loss": 0.593,
+      "step": 117
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 0.35790400007793166,
+      "learning_rate": 4.963937499314857e-05,
+      "loss": 0.6035,
+      "step": 118
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 0.31118450185510343,
+      "learning_rate": 4.962019382530521e-05,
+      "loss": 0.5811,
+      "step": 119
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.3326176465041298,
+      "learning_rate": 4.960051957873725e-05,
+      "loss": 0.581,
+      "step": 120
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 0.30210249377153575,
+      "learning_rate": 4.958035264746893e-05,
+      "loss": 0.5837,
+      "step": 121
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 0.3480385124671555,
+      "learning_rate": 4.955969343539162e-05,
+      "loss": 0.5768,
+      "step": 122
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 0.3003392569743352,
+      "learning_rate": 4.9538542356255866e-05,
+      "loss": 0.5938,
+      "step": 123
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 0.32082565179488104,
+      "learning_rate": 4.9516899833663e-05,
+      "loss": 0.5948,
+      "step": 124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3564349708048278,
+      "learning_rate": 4.949476630105669e-05,
+      "loss": 0.595,
+      "step": 125
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 0.32049541972124757,
+      "learning_rate": 4.94721422017143e-05,
+      "loss": 0.5838,
+      "step": 126
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 0.3317680882353993,
+      "learning_rate": 4.944902798873794e-05,
+      "loss": 0.5952,
+      "step": 127
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 0.3381465061198974,
+      "learning_rate": 4.942542412504543e-05,
+      "loss": 0.6004,
+      "step": 128
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.38351657127693595,
+      "learning_rate": 4.940133108336105e-05,
+      "loss": 0.6014,
+      "step": 129
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 0.3276142738951724,
+      "learning_rate": 4.9376749346206006e-05,
+      "loss": 0.5853,
+      "step": 130
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 0.37146400882939534,
+      "learning_rate": 4.935167940588887e-05,
+      "loss": 0.5995,
+      "step": 131
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.32804274509201087,
+      "learning_rate": 4.9326121764495596e-05,
+      "loss": 0.5955,
+      "step": 132
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 0.3344845806030499,
+      "learning_rate": 4.9300076933879574e-05,
+      "loss": 0.5818,
+      "step": 133
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 0.3479572078392269,
+      "learning_rate": 4.92735454356513e-05,
+      "loss": 0.5941,
+      "step": 134
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.34868252062960353,
+      "learning_rate": 4.924652780116799e-05,
+      "loss": 0.5898,
+      "step": 135
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 0.35674279058993497,
+      "learning_rate": 4.921902457152289e-05,
+      "loss": 0.5899,
+      "step": 136
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 0.3672614416380493,
+      "learning_rate": 4.9191036297534454e-05,
+      "loss": 0.585,
+      "step": 137
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 0.4039478601084677,
+      "learning_rate": 4.916256353973535e-05,
+      "loss": 0.5994,
+      "step": 138
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 0.3428958061155067,
+      "learning_rate": 4.913360686836117e-05,
+      "loss": 0.575,
+      "step": 139
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.4024960602256603,
+      "learning_rate": 4.910416686333906e-05,
+      "loss": 0.5913,
+      "step": 140
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.31040065034832104,
+      "learning_rate": 4.907424411427608e-05,
+      "loss": 0.5761,
+      "step": 141
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 0.359237099401051,
+      "learning_rate": 4.90438392204474e-05,
+      "loss": 0.5885,
+      "step": 142
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 0.3357545415879296,
+      "learning_rate": 4.901295279078431e-05,
+      "loss": 0.5907,
+      "step": 143
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.2846403022642179,
+      "learning_rate": 4.898158544386201e-05,
+      "loss": 0.5886,
+      "step": 144
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.3636245125193307,
+      "learning_rate": 4.894973780788722e-05,
+      "loss": 0.5816,
+      "step": 145
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 0.25440894793562924,
+      "learning_rate": 4.8917410520685635e-05,
+      "loss": 0.576,
+      "step": 146
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 0.3380189678855273,
+      "learning_rate": 4.888460422968908e-05,
+      "loss": 0.5931,
+      "step": 147
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 0.3096794617975588,
+      "learning_rate": 4.885131959192262e-05,
+      "loss": 0.5829,
+      "step": 148
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 0.280174710159943,
+      "learning_rate": 4.881755727399134e-05,
+      "loss": 0.5794,
+      "step": 149
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.31769340776297994,
+      "learning_rate": 4.878331795206705e-05,
+      "loss": 0.5729,
+      "step": 150
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 0.31671973855902796,
+      "learning_rate": 4.8748602311874694e-05,
+      "loss": 0.5905,
+      "step": 151
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 0.32614211009906474,
+      "learning_rate": 4.8713411048678635e-05,
+      "loss": 0.5855,
+      "step": 152
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.29921149443441614,
+      "learning_rate": 4.8677744867268764e-05,
+      "loss": 0.5779,
+      "step": 153
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 0.3558339409344647,
+      "learning_rate": 4.8641604481946314e-05,
+      "loss": 0.5892,
+      "step": 154
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 0.285079025062,
+      "learning_rate": 4.8604990616509616e-05,
+      "loss": 0.5912,
+      "step": 155
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 0.32189736402098207,
+      "learning_rate": 4.856790400423958e-05,
+      "loss": 0.5881,
+      "step": 156
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 0.3293153125716864,
+      "learning_rate": 4.8530345387885004e-05,
+      "loss": 0.5679,
+      "step": 157
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 0.3020107126026594,
+      "learning_rate": 4.849231551964771e-05,
+      "loss": 0.5892,
+      "step": 158
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 0.37859431920215386,
+      "learning_rate": 4.845381516116748e-05,
+      "loss": 0.5752,
+      "step": 159
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.3310159661704287,
+      "learning_rate": 4.841484508350679e-05,
+      "loss": 0.5746,
+      "step": 160
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 0.29255683283621947,
+      "learning_rate": 4.837540606713538e-05,
+      "loss": 0.5822,
+      "step": 161
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 0.35293526712063267,
+      "learning_rate": 4.83354989019146e-05,
+      "loss": 0.5872,
+      "step": 162
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 0.31406471157688987,
+      "learning_rate": 4.829512438708163e-05,
+      "loss": 0.5837,
+      "step": 163
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 0.3746042872826804,
+      "learning_rate": 4.8254283331233464e-05,
+      "loss": 0.5795,
+      "step": 164
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.36631879902112746,
+      "learning_rate": 4.821297655231066e-05,
+      "loss": 0.5917,
+      "step": 165
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 0.31527709590813463,
+      "learning_rate": 4.817120487758104e-05,
+      "loss": 0.583,
+      "step": 166
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 0.3311399262333245,
+      "learning_rate": 4.812896914362309e-05,
+      "loss": 0.5806,
+      "step": 167
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.2797000752887155,
+      "learning_rate": 4.808627019630917e-05,
+      "loss": 0.5818,
+      "step": 168
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 0.34099518073149715,
+      "learning_rate": 4.804310889078861e-05,
+      "loss": 0.5827,
+      "step": 169
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.2858567724381781,
+      "learning_rate": 4.799948609147061e-05,
+      "loss": 0.574,
+      "step": 170
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.3714847802025429,
+      "learning_rate": 4.7955402672006854e-05,
+      "loss": 0.568,
+      "step": 171
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 0.3629775316062746,
+      "learning_rate": 4.791085951527408e-05,
+      "loss": 0.5908,
+      "step": 172
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 0.3308553993376432,
+      "learning_rate": 4.786585751335637e-05,
+      "loss": 0.5822,
+      "step": 173
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.3583796747287126,
+      "learning_rate": 4.782039756752727e-05,
+      "loss": 0.5849,
+      "step": 174
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.3200425393735649,
+      "learning_rate": 4.777448058823179e-05,
+      "loss": 0.5975,
+      "step": 175
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 0.292626050602806,
+      "learning_rate": 4.77281074950681e-05,
+      "loss": 0.5858,
+      "step": 176
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 0.35199899958158143,
+      "learning_rate": 4.768127921676916e-05,
+      "loss": 0.5956,
+      "step": 177
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 0.3049078150389153,
+      "learning_rate": 4.763399669118414e-05,
+      "loss": 0.5818,
+      "step": 178
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 0.3171759989327936,
+      "learning_rate": 4.758626086525956e-05,
+      "loss": 0.5788,
+      "step": 179
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.37616755806228974,
+      "learning_rate": 4.753807269502041e-05,
+      "loss": 0.5764,
+      "step": 180
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 0.3111683206800814,
+      "learning_rate": 4.748943314555093e-05,
+      "loss": 0.5904,
+      "step": 181
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 0.33324357602957505,
+      "learning_rate": 4.744034319097535e-05,
+      "loss": 0.5837,
+      "step": 182
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 0.3033759415187109,
+      "learning_rate": 4.739080381443834e-05,
+      "loss": 0.5872,
+      "step": 183
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 0.31221903432674414,
+      "learning_rate": 4.734081600808531e-05,
+      "loss": 0.5788,
+      "step": 184
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 0.3521014284242888,
+      "learning_rate": 4.7290380773042575e-05,
+      "loss": 0.5876,
+      "step": 185
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 0.310657917751192,
+      "learning_rate": 4.723949911939728e-05,
+      "loss": 0.5799,
+      "step": 186
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 0.3571637318567018,
+      "learning_rate": 4.718817206617718e-05,
+      "loss": 0.5768,
+      "step": 187
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 0.3019122458668047,
+      "learning_rate": 4.713640064133025e-05,
+      "loss": 0.569,
+      "step": 188
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 0.3867595871931151,
+      "learning_rate": 4.7084185881704037e-05,
+      "loss": 0.5807,
+      "step": 189
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.31887858233709804,
+      "learning_rate": 4.7031528833024976e-05,
+      "loss": 0.5853,
+      "step": 190
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 0.33241965371410676,
+      "learning_rate": 4.697843054987737e-05,
+      "loss": 0.5843,
+      "step": 191
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.3186271240585135,
+      "learning_rate": 4.692489209568234e-05,
+      "loss": 0.58,
+      "step": 192
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 0.27257060138401595,
+      "learning_rate": 4.687091454267646e-05,
+      "loss": 0.5782,
+      "step": 193
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 0.3068678257607851,
+      "learning_rate": 4.681649897189036e-05,
+      "loss": 0.5715,
+      "step": 194
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.2895553434845913,
+      "learning_rate": 4.6761646473126985e-05,
+      "loss": 0.5864,
+      "step": 195
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 0.3320177973067205,
+      "learning_rate": 4.670635814493984e-05,
+      "loss": 0.5895,
+      "step": 196
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 0.28847585573804785,
+      "learning_rate": 4.665063509461097e-05,
+      "loss": 0.5906,
+      "step": 197
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.36087996937741346,
+      "learning_rate": 4.6594478438128757e-05,
+      "loss": 0.5827,
+      "step": 198
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 0.2999930376743691,
+      "learning_rate": 4.653788930016562e-05,
+      "loss": 0.5661,
+      "step": 199
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.3488298254178776,
+      "learning_rate": 4.6480868814055424e-05,
+      "loss": 0.5793,
+      "step": 200
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.3489687054350795,
+      "learning_rate": 4.6423418121770855e-05,
+      "loss": 0.5804,
+      "step": 201
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 0.32040065196445794,
+      "learning_rate": 4.636553837390051e-05,
+      "loss": 0.5885,
+      "step": 202
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 0.3211301436932052,
+      "learning_rate": 4.630723072962584e-05,
+      "loss": 0.5712,
+      "step": 203
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.359318175552222,
+      "learning_rate": 4.6248496356697966e-05,
+      "loss": 0.5708,
+      "step": 204
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.28862834349726524,
+      "learning_rate": 4.618933643141428e-05,
+      "loss": 0.5761,
+      "step": 205
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 0.37248048616519797,
+      "learning_rate": 4.6129752138594874e-05,
+      "loss": 0.5853,
+      "step": 206
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.26505735554799875,
+      "learning_rate": 4.6069744671558835e-05,
+      "loss": 0.5711,
+      "step": 207
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 0.4313554228926337,
+      "learning_rate": 4.6009315232100324e-05,
+      "loss": 0.5805,
+      "step": 208
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 0.25367772845462405,
+      "learning_rate": 4.5948465030464536e-05,
+      "loss": 0.5804,
+      "step": 209
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.4241960582103298,
+      "learning_rate": 4.588719528532342e-05,
+      "loss": 0.5781,
+      "step": 210
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 0.26959374978777695,
+      "learning_rate": 4.58255072237513e-05,
+      "loss": 0.5872,
+      "step": 211
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 0.32686893212462864,
+      "learning_rate": 4.5763402081200294e-05,
+      "loss": 0.5865,
+      "step": 212
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.23434165675561655,
+      "learning_rate": 4.570088110147559e-05,
+      "loss": 0.5605,
+      "step": 213
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 0.29221024109189686,
+      "learning_rate": 4.56379455367105e-05,
+      "loss": 0.5798,
+      "step": 214
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 0.24948843663318818,
+      "learning_rate": 4.557459664734141e-05,
+      "loss": 0.5647,
+      "step": 215
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.2697506211061829,
+      "learning_rate": 4.551083570208252e-05,
+      "loss": 0.574,
+      "step": 216
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 0.26982678855787623,
+      "learning_rate": 4.544666397790043e-05,
+      "loss": 0.5769,
+      "step": 217
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 0.26312816652400706,
+      "learning_rate": 4.538208275998861e-05,
+      "loss": 0.5794,
+      "step": 218
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 0.23961265146671684,
+      "learning_rate": 4.531709334174161e-05,
+      "loss": 0.5713,
+      "step": 219
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.2521656675623654,
+      "learning_rate": 4.5251697024729165e-05,
+      "loss": 0.5799,
+      "step": 220
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 0.2554557690482637,
+      "learning_rate": 4.518589511867017e-05,
+      "loss": 0.5612,
+      "step": 221
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.2575994504688328,
+      "learning_rate": 4.511968894140639e-05,
+      "loss": 0.5646,
+      "step": 222
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 0.24148588167926954,
+      "learning_rate": 4.50530798188761e-05,
+      "loss": 0.5752,
+      "step": 223
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 0.24795970441926052,
+      "learning_rate": 4.498606908508754e-05,
+      "loss": 0.5651,
+      "step": 224
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.2594616590711217,
+      "learning_rate": 4.491865808209215e-05,
+      "loss": 0.5677,
+      "step": 225
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 0.25674541102342535,
+      "learning_rate": 4.485084815995778e-05,
+      "loss": 0.5612,
+      "step": 226
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 0.24985240413836018,
+      "learning_rate": 4.478264067674155e-05,
+      "loss": 0.5699,
+      "step": 227
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 0.2565786888027258,
+      "learning_rate": 4.471403699846272e-05,
+      "loss": 0.5649,
+      "step": 228
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 0.24372178698061564,
+      "learning_rate": 4.4645038499075296e-05,
+      "loss": 0.5727,
+      "step": 229
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.26957885657754116,
+      "learning_rate": 4.457564656044056e-05,
+      "loss": 0.5673,
+      "step": 230
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.2607171265747611,
+      "learning_rate": 4.4505862572299315e-05,
+      "loss": 0.5745,
+      "step": 231
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 0.24863876648291447,
+      "learning_rate": 4.443568793224415e-05,
+      "loss": 0.5681,
+      "step": 232
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 0.263062932590105,
+      "learning_rate": 4.436512404569136e-05,
+      "loss": 0.5644,
+      "step": 233
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.2414635376635183,
+      "learning_rate": 4.429417232585288e-05,
+      "loss": 0.5701,
+      "step": 234
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 0.25798244731348213,
+      "learning_rate": 4.422283419370789e-05,
+      "loss": 0.5749,
+      "step": 235
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 0.2705236848186794,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 0.5722,
+      "step": 236
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.2532310162226822,
+      "learning_rate": 4.407900441508084e-05,
+      "loss": 0.5613,
+      "step": 237
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 0.27015507001673206,
+      "learning_rate": 4.400651564913676e-05,
+      "loss": 0.5764,
+      "step": 238
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 0.2600356410106743,
+      "learning_rate": 4.3933646231904504e-05,
+      "loss": 0.5621,
+      "step": 239
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.2947154143151425,
+      "learning_rate": 4.3860397622769756e-05,
+      "loss": 0.5776,
+      "step": 240
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 0.278006374370773,
+      "learning_rate": 4.37867712887125e-05,
+      "loss": 0.5779,
+      "step": 241
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 0.2737057808752138,
+      "learning_rate": 4.371276870427753e-05,
+      "loss": 0.5691,
+      "step": 242
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.2779579791922396,
+      "learning_rate": 4.363839135154497e-05,
+      "loss": 0.5561,
+      "step": 243
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 0.2900125824723869,
+      "learning_rate": 4.356364072010059e-05,
+      "loss": 0.5862,
+      "step": 244
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.2864520071161508,
+      "learning_rate": 4.348851830700593e-05,
+      "loss": 0.5622,
+      "step": 245
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.26378798903912654,
+      "learning_rate": 4.3413025616768424e-05,
+      "loss": 0.5709,
+      "step": 246
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 0.3060187732302713,
+      "learning_rate": 4.333716416131115e-05,
+      "loss": 0.5669,
+      "step": 247
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 0.29432171614697805,
+      "learning_rate": 4.3260935459942584e-05,
+      "loss": 0.5682,
+      "step": 248
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.26017905775079875,
+      "learning_rate": 4.318434103932622e-05,
+      "loss": 0.5885,
+      "step": 249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31505146670137313,
+      "learning_rate": 4.310738243344996e-05,
+      "loss": 0.5761,
+      "step": 250
+    },
+    {
+      "epoch": 0.32128,
+      "grad_norm": 0.28372810809515076,
+      "learning_rate": 4.303006118359537e-05,
+      "loss": 0.5592,
+      "step": 251
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 0.2651991449577939,
+      "learning_rate": 4.295237883830685e-05,
+      "loss": 0.5697,
+      "step": 252
+    },
+    {
+      "epoch": 0.32384,
+      "grad_norm": 0.24913000845932673,
+      "learning_rate": 4.2874336953360616e-05,
+      "loss": 0.5776,
+      "step": 253
+    },
+    {
+      "epoch": 0.32512,
+      "grad_norm": 0.2430027683011546,
+      "learning_rate": 4.2795937091733515e-05,
+      "loss": 0.5677,
+      "step": 254
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.2568626552616791,
+      "learning_rate": 4.271718082357175e-05,
+      "loss": 0.582,
+      "step": 255
+    },
+    {
+      "epoch": 0.32768,
+      "grad_norm": 0.23741828351399039,
+      "learning_rate": 4.2638069726159424e-05,
+      "loss": 0.5609,
+      "step": 256
+    },
+    {
+      "epoch": 0.32896,
+      "grad_norm": 0.23355776394833105,
+      "learning_rate": 4.255860538388694e-05,
+      "loss": 0.572,
+      "step": 257
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 0.28500006430456526,
+      "learning_rate": 4.247878938821929e-05,
+      "loss": 0.5783,
+      "step": 258
+    },
+    {
+      "epoch": 0.33152,
+      "grad_norm": 0.23832612238075213,
+      "learning_rate": 4.2398623337664176e-05,
+      "loss": 0.5591,
+      "step": 259
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.3027829669410571,
+      "learning_rate": 4.231810883773999e-05,
+      "loss": 0.5723,
+      "step": 260
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 0.24742812416416485,
+      "learning_rate": 4.223724750094366e-05,
+      "loss": 0.562,
+      "step": 261
+    },
+    {
+      "epoch": 0.33536,
+      "grad_norm": 0.281426189909417,
+      "learning_rate": 4.215604094671835e-05,
+      "loss": 0.5711,
+      "step": 262
+    },
+    {
+      "epoch": 0.33664,
+      "grad_norm": 0.3464883424573956,
+      "learning_rate": 4.207449080142104e-05,
+      "loss": 0.56,
+      "step": 263
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 0.23536130124235047,
+      "learning_rate": 4.199259869828998e-05,
+      "loss": 0.5714,
+      "step": 264
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 0.32901239458674364,
+      "learning_rate": 4.191036627741191e-05,
+      "loss": 0.5656,
+      "step": 265
+    },
+    {
+      "epoch": 0.34048,
+      "grad_norm": 0.2513197034442368,
+      "learning_rate": 4.182779518568926e-05,
+      "loss": 0.5549,
+      "step": 266
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 0.2956698427587275,
+      "learning_rate": 4.174488707680717e-05,
+      "loss": 0.565,
+      "step": 267
+    },
+    {
+      "epoch": 0.34304,
+      "grad_norm": 0.2767183631207283,
+      "learning_rate": 4.1661643611200366e-05,
+      "loss": 0.5751,
+      "step": 268
+    },
+    {
+      "epoch": 0.34432,
+      "grad_norm": 0.24523313623813994,
+      "learning_rate": 4.157806645601988e-05,
+      "loss": 0.5742,
+      "step": 269
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.2908460538863046,
+      "learning_rate": 4.149415728509971e-05,
+      "loss": 0.5593,
+      "step": 270
+    },
+    {
+      "epoch": 0.34688,
+      "grad_norm": 0.4424923362280487,
+      "learning_rate": 4.140991777892324e-05,
+      "loss": 0.5814,
+      "step": 271
+    },
+    {
+      "epoch": 0.34816,
+      "grad_norm": 0.2575686169084144,
+      "learning_rate": 4.132534962458962e-05,
+      "loss": 0.5678,
+      "step": 272
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 0.2629805233702073,
+      "learning_rate": 4.124045451578001e-05,
+      "loss": 0.5629,
+      "step": 273
+    },
+    {
+      "epoch": 0.35072,
+      "grad_norm": 0.22898476571255702,
+      "learning_rate": 4.115523415272358e-05,
+      "loss": 0.569,
+      "step": 274
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.2443223155892946,
+      "learning_rate": 4.1069690242163484e-05,
+      "loss": 0.5624,
+      "step": 275
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 0.24146954716830435,
+      "learning_rate": 4.0983824497322755e-05,
+      "loss": 0.5783,
+      "step": 276
+    },
+    {
+      "epoch": 0.35456,
+      "grad_norm": 0.23532353314194146,
+      "learning_rate": 4.0897638637869874e-05,
+      "loss": 0.5846,
+      "step": 277
+    },
+    {
+      "epoch": 0.35584,
+      "grad_norm": 0.21666728547483338,
+      "learning_rate": 4.0811134389884433e-05,
+      "loss": 0.5788,
+      "step": 278
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 0.2296647805474214,
+      "learning_rate": 4.07243134858225e-05,
+      "loss": 0.5679,
+      "step": 279
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.23340342966367908,
+      "learning_rate": 4.063717766448194e-05,
+      "loss": 0.5645,
+      "step": 280
+    },
+    {
+      "epoch": 0.35968,
+      "grad_norm": 0.23757239204061076,
+      "learning_rate": 4.05497286709676e-05,
+      "loss": 0.5745,
+      "step": 281
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 0.24284616004111326,
+      "learning_rate": 4.0461968256656376e-05,
+      "loss": 0.5659,
+      "step": 282
+    },
+    {
+      "epoch": 0.36224,
+      "grad_norm": 0.2317609526061923,
+      "learning_rate": 4.037389817916208e-05,
+      "loss": 0.5657,
+      "step": 283
+    },
+    {
+      "epoch": 0.36352,
+      "grad_norm": 0.2779219267379158,
+      "learning_rate": 4.028552020230031e-05,
+      "loss": 0.5612,
+      "step": 284
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.2568685752720875,
+      "learning_rate": 4.019683609605305e-05,
+      "loss": 0.5618,
+      "step": 285
+    },
+    {
+      "epoch": 0.36608,
+      "grad_norm": 0.24357977645804754,
+      "learning_rate": 4.010784763653331e-05,
+      "loss": 0.5532,
+      "step": 286
+    },
+    {
+      "epoch": 0.36736,
+      "grad_norm": 0.2854087705449231,
+      "learning_rate": 4.001855660594948e-05,
+      "loss": 0.5598,
+      "step": 287
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 0.27153778848296095,
+      "learning_rate": 3.9928964792569655e-05,
+      "loss": 0.5661,
+      "step": 288
+    },
+    {
+      "epoch": 0.36992,
+      "grad_norm": 0.3013570035330345,
+      "learning_rate": 3.983907399068587e-05,
+      "loss": 0.576,
+      "step": 289
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.23579225525517492,
+      "learning_rate": 3.974888600057808e-05,
+      "loss": 0.5685,
+      "step": 290
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 0.29344672148557993,
+      "learning_rate": 3.965840262847818e-05,
+      "loss": 0.5652,
+      "step": 291
+    },
+    {
+      "epoch": 0.37376,
+      "grad_norm": 0.2771114973156099,
+      "learning_rate": 3.956762568653378e-05,
+      "loss": 0.5615,
+      "step": 292
+    },
+    {
+      "epoch": 0.37504,
+      "grad_norm": 0.2475727777258234,
+      "learning_rate": 3.947655699277197e-05,
+      "loss": 0.5622,
+      "step": 293
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 0.3283313482469374,
+      "learning_rate": 3.9385198371062845e-05,
+      "loss": 0.5688,
+      "step": 294
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 0.2567128249078116,
+      "learning_rate": 3.929355165108299e-05,
+      "loss": 0.5685,
+      "step": 295
+    },
+    {
+      "epoch": 0.37888,
+      "grad_norm": 0.33868790476694477,
+      "learning_rate": 3.920161866827889e-05,
+      "loss": 0.5603,
+      "step": 296
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 0.23159577247265578,
+      "learning_rate": 3.910940126383013e-05,
+      "loss": 0.5614,
+      "step": 297
+    },
+    {
+      "epoch": 0.38144,
+      "grad_norm": 0.261651084842406,
+      "learning_rate": 3.9016901284612474e-05,
+      "loss": 0.5634,
+      "step": 298
+    },
+    {
+      "epoch": 0.38272,
+      "grad_norm": 0.25918417083311507,
+      "learning_rate": 3.8924120583160985e-05,
+      "loss": 0.564,
+      "step": 299
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.24530554345595257,
+      "learning_rate": 3.883106101763285e-05,
+      "loss": 0.5751,
+      "step": 300
+    },
+    {
+      "epoch": 0.38528,
+      "grad_norm": 0.25269762220161,
+      "learning_rate": 3.873772445177015e-05,
+      "loss": 0.5666,
+      "step": 301
+    },
+    {
+      "epoch": 0.38656,
+      "grad_norm": 0.2686509936643206,
+      "learning_rate": 3.8644112754862614e-05,
+      "loss": 0.5787,
+      "step": 302
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 0.2306788708136305,
+      "learning_rate": 3.85502278017101e-05,
+      "loss": 0.5639,
+      "step": 303
+    },
+    {
+      "epoch": 0.38912,
+      "grad_norm": 0.25921338779127046,
+      "learning_rate": 3.84560714725851e-05,
+      "loss": 0.5611,
+      "step": 304
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 0.2875441041443726,
+      "learning_rate": 3.8361645653195026e-05,
+      "loss": 0.5641,
+      "step": 305
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 0.24679004673798802,
+      "learning_rate": 3.8266952234644545e-05,
+      "loss": 0.5696,
+      "step": 306
+    },
+    {
+      "epoch": 0.39296,
+      "grad_norm": 0.32137055545965876,
+      "learning_rate": 3.817199311339759e-05,
+      "loss": 0.5642,
+      "step": 307
+    },
+    {
+      "epoch": 0.39424,
+      "grad_norm": 0.25511061251478745,
+      "learning_rate": 3.807677019123944e-05,
+      "loss": 0.5535,
+      "step": 308
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 0.28918335928655264,
+      "learning_rate": 3.798128537523865e-05,
+      "loss": 0.5628,
+      "step": 309
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.24654149918381577,
+      "learning_rate": 3.7885540577708804e-05,
+      "loss": 0.5693,
+      "step": 310
+    },
+    {
+      "epoch": 0.39808,
+      "grad_norm": 0.29445546859209903,
+      "learning_rate": 3.7789537716170256e-05,
+      "loss": 0.5603,
+      "step": 311
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 0.2546352881022481,
+      "learning_rate": 3.76932787133117e-05,
+      "loss": 0.5605,
+      "step": 312
+    },
+    {
+      "epoch": 0.40064,
+      "grad_norm": 0.2694209869990075,
+      "learning_rate": 3.759676549695168e-05,
+      "loss": 0.5631,
+      "step": 313
+    },
+    {
+      "epoch": 0.40192,
+      "grad_norm": 0.25374147038521105,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.5737,
+      "step": 314
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.2692940305682822,
+      "learning_rate": 3.740298416041898e-05,
+      "loss": 0.5592,
+      "step": 315
+    },
+    {
+      "epoch": 0.40448,
+      "grad_norm": 0.27292894144562746,
+      "learning_rate": 3.730571992118462e-05,
+      "loss": 0.56,
+      "step": 316
+    },
+    {
+      "epoch": 0.40576,
+      "grad_norm": 0.2831296008226712,
+      "learning_rate": 3.720820923024778e-05,
+      "loss": 0.556,
+      "step": 317
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 0.24842153386433,
+      "learning_rate": 3.711045404049507e-05,
+      "loss": 0.5615,
+      "step": 318
+    },
+    {
+      "epoch": 0.40832,
+      "grad_norm": 0.25137316651759817,
+      "learning_rate": 3.701245630970979e-05,
+      "loss": 0.5519,
+      "step": 319
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.256842976111285,
+      "learning_rate": 3.69142180005327e-05,
+      "loss": 0.559,
+      "step": 320
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 0.23856265603702867,
+      "learning_rate": 3.681574108042274e-05,
+      "loss": 0.5513,
+      "step": 321
+    },
+    {
+      "epoch": 0.41216,
+      "grad_norm": 0.2636505871798519,
+      "learning_rate": 3.6717027521617595e-05,
+      "loss": 0.5553,
+      "step": 322
+    },
+    {
+      "epoch": 0.41344,
+      "grad_norm": 0.23126506975873334,
+      "learning_rate": 3.6618079301094216e-05,
+      "loss": 0.5535,
+      "step": 323
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 0.2765627545708339,
+      "learning_rate": 3.6518898400529214e-05,
+      "loss": 0.5727,
+      "step": 324
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.25065312946741247,
+      "learning_rate": 3.6419486806259194e-05,
+      "loss": 0.5642,
+      "step": 325
+    },
+    {
+      "epoch": 0.41728,
+      "grad_norm": 0.24801934318992586,
+      "learning_rate": 3.631984650924094e-05,
+      "loss": 0.5489,
+      "step": 326
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 0.2505101313201257,
+      "learning_rate": 3.621997950501156e-05,
+      "loss": 0.5575,
+      "step": 327
+    },
+    {
+      "epoch": 0.41984,
+      "grad_norm": 0.22423685749771013,
+      "learning_rate": 3.611988779364853e-05,
+      "loss": 0.5468,
+      "step": 328
+    },
+    {
+      "epoch": 0.42112,
+      "grad_norm": 0.2613413831880175,
+      "learning_rate": 3.6019573379729643e-05,
+      "loss": 0.5742,
+      "step": 329
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.23270975616366224,
+      "learning_rate": 3.591903827229282e-05,
+      "loss": 0.567,
+      "step": 330
+    },
+    {
+      "epoch": 0.42368,
+      "grad_norm": 0.22747265941762546,
+      "learning_rate": 3.5818284484795904e-05,
+      "loss": 0.5521,
+      "step": 331
+    },
+    {
+      "epoch": 0.42496,
+      "grad_norm": 0.21305915578050127,
+      "learning_rate": 3.5717314035076355e-05,
+      "loss": 0.5507,
+      "step": 332
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 0.22564541882328742,
+      "learning_rate": 3.56161289453108e-05,
+      "loss": 0.5612,
+      "step": 333
+    },
+    {
+      "epoch": 0.42752,
+      "grad_norm": 0.24927047862972154,
+      "learning_rate": 3.5514731241974544e-05,
+      "loss": 0.5511,
+      "step": 334
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.23391865199996986,
+      "learning_rate": 3.5413122955801005e-05,
+      "loss": 0.5638,
+      "step": 335
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 0.2538820004383474,
+      "learning_rate": 3.5311306121741015e-05,
+      "loss": 0.5629,
+      "step": 336
+    },
+    {
+      "epoch": 0.43136,
+      "grad_norm": 0.23071027016549392,
+      "learning_rate": 3.5209282778922106e-05,
+      "loss": 0.5524,
+      "step": 337
+    },
+    {
+      "epoch": 0.43264,
+      "grad_norm": 0.2538902369228176,
+      "learning_rate": 3.510705497060762e-05,
+      "loss": 0.5589,
+      "step": 338
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 0.2396787654237747,
+      "learning_rate": 3.500462474415584e-05,
+      "loss": 0.5583,
+      "step": 339
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.24498876339943568,
+      "learning_rate": 3.490199415097892e-05,
+      "loss": 0.5586,
+      "step": 340
+    },
+    {
+      "epoch": 0.43648,
+      "grad_norm": 0.22429707447384664,
+      "learning_rate": 3.479916524650188e-05,
+      "loss": 0.5591,
+      "step": 341
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 0.2488615974495946,
+      "learning_rate": 3.4696140090121376e-05,
+      "loss": 0.5716,
+      "step": 342
+    },
+    {
+      "epoch": 0.43904,
+      "grad_norm": 0.24601068839661924,
+      "learning_rate": 3.459292074516449e-05,
+      "loss": 0.572,
+      "step": 343
+    },
+    {
+      "epoch": 0.44032,
+      "grad_norm": 0.24772654051656678,
+      "learning_rate": 3.4489509278847414e-05,
+      "loss": 0.5627,
+      "step": 344
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.24320980786790716,
+      "learning_rate": 3.4385907762234e-05,
+      "loss": 0.556,
+      "step": 345
+    },
+    {
+      "epoch": 0.44288,
+      "grad_norm": 0.2899880667030433,
+      "learning_rate": 3.428211827019434e-05,
+      "loss": 0.5625,
+      "step": 346
+    },
+    {
+      "epoch": 0.44416,
+      "grad_norm": 0.24951619342805165,
+      "learning_rate": 3.417814288136319e-05,
+      "loss": 0.5606,
+      "step": 347
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 0.29645948538693856,
+      "learning_rate": 3.407398367809832e-05,
+      "loss": 0.5696,
+      "step": 348
+    },
+    {
+      "epoch": 0.44672,
+      "grad_norm": 0.2628573817303425,
+      "learning_rate": 3.3969642746438836e-05,
+      "loss": 0.5548,
+      "step": 349
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.25156314937857505,
+      "learning_rate": 3.386512217606339e-05,
+      "loss": 0.5584,
+      "step": 350
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 0.24416044960818195,
+      "learning_rate": 3.3760424060248344e-05,
+      "loss": 0.5501,
+      "step": 351
+    },
+    {
+      "epoch": 0.45056,
+      "grad_norm": 0.2508322157210064,
+      "learning_rate": 3.365555049582582e-05,
+      "loss": 0.5575,
+      "step": 352
+    },
+    {
+      "epoch": 0.45184,
+      "grad_norm": 0.23612772269109591,
+      "learning_rate": 3.355050358314172e-05,
+      "loss": 0.5682,
+      "step": 353
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 0.24541798896943803,
+      "learning_rate": 3.3445285426013685e-05,
+      "loss": 0.551,
+      "step": 354
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 0.2234016182436274,
+      "learning_rate": 3.3339898131688914e-05,
+      "loss": 0.5591,
+      "step": 355
+    },
+    {
+      "epoch": 0.45568,
+      "grad_norm": 0.24455388410655837,
+      "learning_rate": 3.323434381080199e-05,
+      "loss": 0.5542,
+      "step": 356
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 0.2447461284103842,
+      "learning_rate": 3.312862457733263e-05,
+      "loss": 0.5689,
+      "step": 357
+    },
+    {
+      "epoch": 0.45824,
+      "grad_norm": 0.2611063437561967,
+      "learning_rate": 3.302274254856329e-05,
+      "loss": 0.5564,
+      "step": 358
+    },
+    {
+      "epoch": 0.45952,
+      "grad_norm": 0.2609646712338443,
+      "learning_rate": 3.2916699845036816e-05,
+      "loss": 0.5609,
+      "step": 359
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.2668094661495938,
+      "learning_rate": 3.281049859051394e-05,
+      "loss": 0.55,
+      "step": 360
+    },
+    {
+      "epoch": 0.46208,
+      "grad_norm": 0.23869480950176758,
+      "learning_rate": 3.270414091193077e-05,
+      "loss": 0.5613,
+      "step": 361
+    },
+    {
+      "epoch": 0.46336,
+      "grad_norm": 0.24021036759397824,
+      "learning_rate": 3.2597628939356175e-05,
+      "loss": 0.5431,
+      "step": 362
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 0.22964918725760114,
+      "learning_rate": 3.2490964805949145e-05,
+      "loss": 0.5568,
+      "step": 363
+    },
+    {
+      "epoch": 0.46592,
+      "grad_norm": 0.22304779121360874,
+      "learning_rate": 3.238415064791603e-05,
+      "loss": 0.561,
+      "step": 364
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 0.23921210910709523,
+      "learning_rate": 3.227718860446782e-05,
+      "loss": 0.5457,
+      "step": 365
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 0.2209452744046842,
+      "learning_rate": 3.217008081777726e-05,
+      "loss": 0.5543,
+      "step": 366
+    },
+    {
+      "epoch": 0.46976,
+      "grad_norm": 0.24033224570485728,
+      "learning_rate": 3.206282943293593e-05,
+      "loss": 0.5491,
+      "step": 367
+    },
+    {
+      "epoch": 0.47104,
+      "grad_norm": 0.23799116455210376,
+      "learning_rate": 3.195543659791132e-05,
+      "loss": 0.5473,
+      "step": 368
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 0.22503904067880973,
+      "learning_rate": 3.1847904463503816e-05,
+      "loss": 0.5633,
+      "step": 369
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.22994992202231845,
+      "learning_rate": 3.17402351833036e-05,
+      "loss": 0.56,
+      "step": 370
+    },
+    {
+      "epoch": 0.47488,
+      "grad_norm": 0.2454490715349298,
+      "learning_rate": 3.163243091364752e-05,
+      "loss": 0.5502,
+      "step": 371
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 0.2567184232746229,
+      "learning_rate": 3.152449381357593e-05,
+      "loss": 0.5606,
+      "step": 372
+    },
+    {
+      "epoch": 0.47744,
+      "grad_norm": 0.22104709877502515,
+      "learning_rate": 3.141642604478942e-05,
+      "loss": 0.5452,
+      "step": 373
+    },
+    {
+      "epoch": 0.47872,
+      "grad_norm": 0.254791340469147,
+      "learning_rate": 3.130822977160554e-05,
+      "loss": 0.5575,
+      "step": 374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2172815217297066,
+      "learning_rate": 3.119990716091546e-05,
+      "loss": 0.5592,
+      "step": 375
+    },
+    {
+      "epoch": 0.48128,
+      "grad_norm": 0.23931030435414483,
+      "learning_rate": 3.109146038214055e-05,
+      "loss": 0.5473,
+      "step": 376
+    },
+    {
+      "epoch": 0.48256,
+      "grad_norm": 0.2290268854716801,
+      "learning_rate": 3.098289160718895e-05,
+      "loss": 0.5645,
+      "step": 377
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 0.23603315462362215,
+      "learning_rate": 3.087420301041206e-05,
+      "loss": 0.5408,
+      "step": 378
+    },
+    {
+      "epoch": 0.48512,
+      "grad_norm": 0.20930260887273594,
+      "learning_rate": 3.076539676856101e-05,
+      "loss": 0.5497,
+      "step": 379
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.20115083009307477,
+      "learning_rate": 3.065647506074306e-05,
+      "loss": 0.565,
+      "step": 380
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 0.24482468983205832,
+      "learning_rate": 3.054744006837794e-05,
+      "loss": 0.5682,
+      "step": 381
+    },
+    {
+      "epoch": 0.48896,
+      "grad_norm": 0.21118406075167703,
+      "learning_rate": 3.0438293975154186e-05,
+      "loss": 0.5477,
+      "step": 382
+    },
+    {
+      "epoch": 0.49024,
+      "grad_norm": 0.23616813058785022,
+      "learning_rate": 3.03290389669854e-05,
+      "loss": 0.5563,
+      "step": 383
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 0.21365522272880177,
+      "learning_rate": 3.021967723196647e-05,
+      "loss": 0.5507,
+      "step": 384
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 0.22727170936570085,
+      "learning_rate": 3.0110210960329725e-05,
+      "loss": 0.5603,
+      "step": 385
+    },
+    {
+      "epoch": 0.49408,
+      "grad_norm": 0.20731397878172675,
+      "learning_rate": 3.0000642344401113e-05,
+      "loss": 0.5563,
+      "step": 386
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 0.21475557099958575,
+      "learning_rate": 2.9890973578556268e-05,
+      "loss": 0.552,
+      "step": 387
+    },
+    {
+      "epoch": 0.49664,
+      "grad_norm": 0.26877099196744664,
+      "learning_rate": 2.978120685917656e-05,
+      "loss": 0.5409,
+      "step": 388
+    },
+    {
+      "epoch": 0.49792,
+      "grad_norm": 0.21789486858704812,
+      "learning_rate": 2.9671344384605127e-05,
+      "loss": 0.5486,
+      "step": 389
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.23650494006372377,
+      "learning_rate": 2.956138835510282e-05,
+      "loss": 0.5454,
+      "step": 390
+    },
+    {
+      "epoch": 0.50048,
+      "grad_norm": 0.22781069793383732,
+      "learning_rate": 2.945134097280417e-05,
+      "loss": 0.5615,
+      "step": 391
+    },
+    {
+      "epoch": 0.50176,
+      "grad_norm": 0.2365389525743412,
+      "learning_rate": 2.9341204441673266e-05,
+      "loss": 0.5544,
+      "step": 392
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 0.21603949855780968,
+      "learning_rate": 2.9230980967459593e-05,
+      "loss": 0.569,
+      "step": 393
+    },
+    {
+      "epoch": 0.50432,
+      "grad_norm": 0.2037144247370564,
+      "learning_rate": 2.9120672757653916e-05,
+      "loss": 0.5497,
+      "step": 394
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.23323915671730888,
+      "learning_rate": 2.9010282021444008e-05,
+      "loss": 0.5583,
+      "step": 395
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 0.21232243650279164,
+      "learning_rate": 2.8899810969670448e-05,
+      "loss": 0.5632,
+      "step": 396
+    },
+    {
+      "epoch": 0.50816,
+      "grad_norm": 0.23946451797290258,
+      "learning_rate": 2.8789261814782316e-05,
+      "loss": 0.5593,
+      "step": 397
+    },
+    {
+      "epoch": 0.50944,
+      "grad_norm": 0.2277720908380823,
+      "learning_rate": 2.8678636770792906e-05,
+      "loss": 0.5557,
+      "step": 398
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 0.2121131449498051,
+      "learning_rate": 2.856793805323536e-05,
+      "loss": 0.5494,
+      "step": 399
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.2257692000346662,
+      "learning_rate": 2.845716787911833e-05,
+      "loss": 0.5669,
+      "step": 400
+    },
+    {
+      "epoch": 0.51328,
+      "grad_norm": 0.23206067761113425,
+      "learning_rate": 2.8346328466881545e-05,
+      "loss": 0.5587,
+      "step": 401
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 0.22901702627476864,
+      "learning_rate": 2.8235422036351382e-05,
+      "loss": 0.5719,
+      "step": 402
+    },
+    {
+      "epoch": 0.51584,
+      "grad_norm": 0.2284208784515934,
+      "learning_rate": 2.812445080869646e-05,
+      "loss": 0.5608,
+      "step": 403
+    },
+    {
+      "epoch": 0.51712,
+      "grad_norm": 0.21851283140724692,
+      "learning_rate": 2.8013417006383076e-05,
+      "loss": 0.5567,
+      "step": 404
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.21275077693373082,
+      "learning_rate": 2.7902322853130757e-05,
+      "loss": 0.545,
+      "step": 405
+    },
+    {
+      "epoch": 0.51968,
+      "grad_norm": 0.2251800050901655,
+      "learning_rate": 2.77911705738677e-05,
+      "loss": 0.555,
+      "step": 406
+    },
+    {
+      "epoch": 0.52096,
+      "grad_norm": 0.21241829251347838,
+      "learning_rate": 2.7679962394686198e-05,
+      "loss": 0.5581,
+      "step": 407
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 0.21703951180685088,
+      "learning_rate": 2.756870054279811e-05,
+      "loss": 0.5488,
+      "step": 408
+    },
+    {
+      "epoch": 0.52352,
+      "grad_norm": 0.22069571037639818,
+      "learning_rate": 2.745738724649018e-05,
+      "loss": 0.5535,
+      "step": 409
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.25932488146409416,
+      "learning_rate": 2.7346024735079486e-05,
+      "loss": 0.5583,
+      "step": 410
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 0.20727956981757625,
+      "learning_rate": 2.7234615238868732e-05,
+      "loss": 0.5632,
+      "step": 411
+    },
+    {
+      "epoch": 0.52736,
+      "grad_norm": 0.24066945021398153,
+      "learning_rate": 2.712316098910162e-05,
+      "loss": 0.563,
+      "step": 412
+    },
+    {
+      "epoch": 0.52864,
+      "grad_norm": 0.25008559453154683,
+      "learning_rate": 2.7011664217918154e-05,
+      "loss": 0.5488,
+      "step": 413
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 0.23013581555834695,
+      "learning_rate": 2.6900127158309903e-05,
+      "loss": 0.5633,
+      "step": 414
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 0.22224479213901044,
+      "learning_rate": 2.6788552044075344e-05,
+      "loss": 0.549,
+      "step": 415
+    },
+    {
+      "epoch": 0.53248,
+      "grad_norm": 0.2654591019527864,
+      "learning_rate": 2.667694110977506e-05,
+      "loss": 0.5574,
+      "step": 416
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 0.19545471307044301,
+      "learning_rate": 2.656529659068705e-05,
+      "loss": 0.5488,
+      "step": 417
+    },
+    {
+      "epoch": 0.53504,
+      "grad_norm": 0.20254148474738726,
+      "learning_rate": 2.6453620722761896e-05,
+      "loss": 0.5638,
+      "step": 418
+    },
+    {
+      "epoch": 0.53632,
+      "grad_norm": 0.22685700497041825,
+      "learning_rate": 2.6341915742578037e-05,
+      "loss": 0.5414,
+      "step": 419
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.20942540533409634,
+      "learning_rate": 2.6230183887296955e-05,
+      "loss": 0.5565,
+      "step": 420
+    },
+    {
+      "epoch": 0.53888,
+      "grad_norm": 0.21464639120298495,
+      "learning_rate": 2.6118427394618357e-05,
+      "loss": 0.5561,
+      "step": 421
+    },
+    {
+      "epoch": 0.54016,
+      "grad_norm": 0.22217802407368148,
+      "learning_rate": 2.600664850273538e-05,
+      "loss": 0.5533,
+      "step": 422
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 0.18275865926888668,
+      "learning_rate": 2.5894849450289764e-05,
+      "loss": 0.5488,
+      "step": 423
+    },
+    {
+      "epoch": 0.54272,
+      "grad_norm": 0.2167014210634381,
+      "learning_rate": 2.5783032476327007e-05,
+      "loss": 0.5461,
+      "step": 424
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.2034252786956083,
+      "learning_rate": 2.5671199820251534e-05,
+      "loss": 0.5509,
+      "step": 425
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 0.18569564751312728,
+      "learning_rate": 2.5559353721781832e-05,
+      "loss": 0.5478,
+      "step": 426
+    },
+    {
+      "epoch": 0.54656,
+      "grad_norm": 0.20022896045042796,
+      "learning_rate": 2.544749642090561e-05,
+      "loss": 0.5477,
+      "step": 427
+    },
+    {
+      "epoch": 0.54784,
+      "grad_norm": 0.21037283126869577,
+      "learning_rate": 2.5335630157834937e-05,
+      "loss": 0.5539,
+      "step": 428
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 0.19972461705925887,
+      "learning_rate": 2.522375717296137e-05,
+      "loss": 0.5569,
+      "step": 429
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.19005063735053182,
+      "learning_rate": 2.5111879706811087e-05,
+      "loss": 0.5457,
+      "step": 430
+    },
+    {
+      "epoch": 0.55168,
+      "grad_norm": 0.21356314651647584,
+      "learning_rate": 2.5e-05,
+      "loss": 0.5459,
+      "step": 431
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 0.20945885149378243,
+      "learning_rate": 2.4888120293188916e-05,
+      "loss": 0.5492,
+      "step": 432
+    },
+    {
+      "epoch": 0.55424,
+      "grad_norm": 0.2230850015397349,
+      "learning_rate": 2.4776242827038636e-05,
+      "loss": 0.5545,
+      "step": 433
+    },
+    {
+      "epoch": 0.55552,
+      "grad_norm": 0.23589759701529497,
+      "learning_rate": 2.4664369842165068e-05,
+      "loss": 0.5412,
+      "step": 434
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.22581333235197817,
+      "learning_rate": 2.4552503579094397e-05,
+      "loss": 0.5435,
+      "step": 435
+    },
+    {
+      "epoch": 0.55808,
+      "grad_norm": 0.24919774658733163,
+      "learning_rate": 2.4440646278218177e-05,
+      "loss": 0.545,
+      "step": 436
+    },
+    {
+      "epoch": 0.55936,
+      "grad_norm": 0.2190295213590674,
+      "learning_rate": 2.4328800179748475e-05,
+      "loss": 0.5408,
+      "step": 437
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 0.2628486805976341,
+      "learning_rate": 2.4216967523673e-05,
+      "loss": 0.5655,
+      "step": 438
+    },
+    {
+      "epoch": 0.56192,
+      "grad_norm": 0.25252172819470975,
+      "learning_rate": 2.4105150549710238e-05,
+      "loss": 0.5552,
+      "step": 439
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.20229952269202467,
+      "learning_rate": 2.399335149726463e-05,
+      "loss": 0.5429,
+      "step": 440
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 0.24469450227466819,
+      "learning_rate": 2.388157260538165e-05,
+      "loss": 0.5582,
+      "step": 441
+    },
+    {
+      "epoch": 0.56576,
+      "grad_norm": 0.21164210405032619,
+      "learning_rate": 2.3769816112703047e-05,
+      "loss": 0.5506,
+      "step": 442
+    },
+    {
+      "epoch": 0.56704,
+      "grad_norm": 0.22950726772051805,
+      "learning_rate": 2.365808425742196e-05,
+      "loss": 0.5497,
+      "step": 443
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 0.23336024035401096,
+      "learning_rate": 2.3546379277238107e-05,
+      "loss": 0.5415,
+      "step": 444
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.19765357154592897,
+      "learning_rate": 2.3434703409312954e-05,
+      "loss": 0.5423,
+      "step": 445
+    },
+    {
+      "epoch": 0.57088,
+      "grad_norm": 0.20108815980408287,
+      "learning_rate": 2.3323058890224938e-05,
+      "loss": 0.5543,
+      "step": 446
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 0.20801363546967222,
+      "learning_rate": 2.321144795592467e-05,
+      "loss": 0.5579,
+      "step": 447
+    },
+    {
+      "epoch": 0.57344,
+      "grad_norm": 0.19116838604162356,
+      "learning_rate": 2.3099872841690103e-05,
+      "loss": 0.5418,
+      "step": 448
+    },
+    {
+      "epoch": 0.57472,
+      "grad_norm": 0.18998557353906567,
+      "learning_rate": 2.2988335782081855e-05,
+      "loss": 0.5575,
+      "step": 449
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.20390602126403667,
+      "learning_rate": 2.2876839010898377e-05,
+      "loss": 0.5517,
+      "step": 450
+    },
+    {
+      "epoch": 0.57728,
+      "grad_norm": 0.20273804136737242,
+      "learning_rate": 2.2765384761131274e-05,
+      "loss": 0.5512,
+      "step": 451
+    },
+    {
+      "epoch": 0.57856,
+      "grad_norm": 0.2323345427375102,
+      "learning_rate": 2.265397526492052e-05,
+      "loss": 0.5552,
+      "step": 452
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 0.21618001641053508,
+      "learning_rate": 2.2542612753509823e-05,
+      "loss": 0.5497,
+      "step": 453
+    },
+    {
+      "epoch": 0.58112,
+      "grad_norm": 0.2089693691856067,
+      "learning_rate": 2.24312994572019e-05,
+      "loss": 0.5437,
+      "step": 454
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.21321327766142864,
+      "learning_rate": 2.2320037605313808e-05,
+      "loss": 0.5457,
+      "step": 455
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 0.1866665167284803,
+      "learning_rate": 2.2208829426132307e-05,
+      "loss": 0.5589,
+      "step": 456
+    },
+    {
+      "epoch": 0.58496,
+      "grad_norm": 0.19549868239805784,
+      "learning_rate": 2.2097677146869242e-05,
+      "loss": 0.5428,
+      "step": 457
+    },
+    {
+      "epoch": 0.58624,
+      "grad_norm": 0.21465688730228774,
+      "learning_rate": 2.1986582993616926e-05,
+      "loss": 0.5409,
+      "step": 458
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 0.19658929521150698,
+      "learning_rate": 2.1875549191303545e-05,
+      "loss": 0.5464,
+      "step": 459
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.198475793396338,
+      "learning_rate": 2.1764577963648614e-05,
+      "loss": 0.5468,
+      "step": 460
+    },
+    {
+      "epoch": 0.59008,
+      "grad_norm": 0.21158010098295732,
+      "learning_rate": 2.1653671533118468e-05,
+      "loss": 0.5537,
+      "step": 461
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 0.21320070068715855,
+      "learning_rate": 2.154283212088168e-05,
+      "loss": 0.5486,
+      "step": 462
+    },
+    {
+      "epoch": 0.59264,
+      "grad_norm": 0.2212774837957101,
+      "learning_rate": 2.1432061946764644e-05,
+      "loss": 0.5381,
+      "step": 463
+    },
+    {
+      "epoch": 0.59392,
+      "grad_norm": 0.19109434465910716,
+      "learning_rate": 2.1321363229207096e-05,
+      "loss": 0.5538,
+      "step": 464
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.194950360975785,
+      "learning_rate": 2.121073818521769e-05,
+      "loss": 0.5539,
+      "step": 465
+    },
+    {
+      "epoch": 0.59648,
+      "grad_norm": 0.20628175723560369,
+      "learning_rate": 2.1100189030329558e-05,
+      "loss": 0.5414,
+      "step": 466
+    },
+    {
+      "epoch": 0.59776,
+      "grad_norm": 0.19027258987716833,
+      "learning_rate": 2.098971797855599e-05,
+      "loss": 0.5465,
+      "step": 467
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 0.1968353186636271,
+      "learning_rate": 2.0879327242346093e-05,
+      "loss": 0.5429,
+      "step": 468
+    },
+    {
+      "epoch": 0.60032,
+      "grad_norm": 0.19197267253975747,
+      "learning_rate": 2.0769019032540416e-05,
+      "loss": 0.5576,
+      "step": 469
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.1989897674210096,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 0.5421,
+      "step": 470
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 0.2081969628582826,
+      "learning_rate": 2.054865902719584e-05,
+      "loss": 0.567,
+      "step": 471
+    },
+    {
+      "epoch": 0.60416,
+      "grad_norm": 0.19431308789862117,
+      "learning_rate": 2.043861164489719e-05,
+      "loss": 0.5659,
+      "step": 472
+    },
+    {
+      "epoch": 0.60544,
+      "grad_norm": 0.21246150041771483,
+      "learning_rate": 2.0328655615394882e-05,
+      "loss": 0.5457,
+      "step": 473
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 0.20631939020361545,
+      "learning_rate": 2.021879314082344e-05,
+      "loss": 0.5461,
+      "step": 474
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.18725048532643526,
+      "learning_rate": 2.0109026421443745e-05,
+      "loss": 0.5451,
+      "step": 475
+    },
+    {
+      "epoch": 0.60928,
+      "grad_norm": 0.20394496506934465,
+      "learning_rate": 1.9999357655598893e-05,
+      "loss": 0.5575,
+      "step": 476
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 0.22485652668287356,
+      "learning_rate": 1.9889789039670277e-05,
+      "loss": 0.553,
+      "step": 477
+    },
+    {
+      "epoch": 0.61184,
+      "grad_norm": 0.20345540078269608,
+      "learning_rate": 1.978032276803354e-05,
+      "loss": 0.55,
+      "step": 478
+    },
+    {
+      "epoch": 0.61312,
+      "grad_norm": 0.2062068875047177,
+      "learning_rate": 1.9670961033014605e-05,
+      "loss": 0.5473,
+      "step": 479
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.18808340518062924,
+      "learning_rate": 1.956170602484582e-05,
+      "loss": 0.5481,
+      "step": 480
+    },
+    {
+      "epoch": 0.61568,
+      "grad_norm": 0.1948648618263434,
+      "learning_rate": 1.9452559931622067e-05,
+      "loss": 0.5611,
+      "step": 481
+    },
+    {
+      "epoch": 0.61696,
+      "grad_norm": 0.21011586889216233,
+      "learning_rate": 1.934352493925695e-05,
+      "loss": 0.558,
+      "step": 482
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 0.21372649563774826,
+      "learning_rate": 1.9234603231438995e-05,
+      "loss": 0.5504,
+      "step": 483
+    },
+    {
+      "epoch": 0.61952,
+      "grad_norm": 0.21202134397012273,
+      "learning_rate": 1.9125796989587947e-05,
+      "loss": 0.5423,
+      "step": 484
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.1813013729625519,
+      "learning_rate": 1.9017108392811065e-05,
+      "loss": 0.5478,
+      "step": 485
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 0.19357846387987115,
+      "learning_rate": 1.8908539617859456e-05,
+      "loss": 0.5371,
+      "step": 486
+    },
+    {
+      "epoch": 0.62336,
+      "grad_norm": 0.22921908791193235,
+      "learning_rate": 1.880009283908454e-05,
+      "loss": 0.5508,
+      "step": 487
+    },
+    {
+      "epoch": 0.62464,
+      "grad_norm": 0.17331042850311895,
+      "learning_rate": 1.8691770228394456e-05,
+      "loss": 0.5494,
+      "step": 488
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 0.19359927189150627,
+      "learning_rate": 1.858357395521058e-05,
+      "loss": 0.5597,
+      "step": 489
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.18648874954378045,
+      "learning_rate": 1.8475506186424074e-05,
+      "loss": 0.5484,
+      "step": 490
+    },
+    {
+      "epoch": 0.62848,
+      "grad_norm": 0.1886503036851335,
+      "learning_rate": 1.8367569086352483e-05,
+      "loss": 0.5654,
+      "step": 491
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 0.17838832879819544,
+      "learning_rate": 1.825976481669641e-05,
+      "loss": 0.5441,
+      "step": 492
+    },
+    {
+      "epoch": 0.63104,
+      "grad_norm": 0.18783084375563594,
+      "learning_rate": 1.815209553649619e-05,
+      "loss": 0.5499,
+      "step": 493
+    },
+    {
+      "epoch": 0.63232,
+      "grad_norm": 0.17314646438183282,
+      "learning_rate": 1.8044563402088684e-05,
+      "loss": 0.5448,
+      "step": 494
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.18733598616464592,
+      "learning_rate": 1.7937170567064075e-05,
+      "loss": 0.5664,
+      "step": 495
+    },
+    {
+      "epoch": 0.63488,
+      "grad_norm": 0.20893112653690207,
+      "learning_rate": 1.7829919182222752e-05,
+      "loss": 0.537,
+      "step": 496
+    },
+    {
+      "epoch": 0.63616,
+      "grad_norm": 0.18275232842737502,
+      "learning_rate": 1.7722811395532178e-05,
+      "loss": 0.556,
+      "step": 497
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 0.1735031563340356,
+      "learning_rate": 1.7615849352083975e-05,
+      "loss": 0.5541,
+      "step": 498
+    },
+    {
+      "epoch": 0.63872,
+      "grad_norm": 0.19136749409935053,
+      "learning_rate": 1.7509035194050868e-05,
+      "loss": 0.5394,
+      "step": 499
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18335262719738063,
+      "learning_rate": 1.740237106064383e-05,
+      "loss": 0.5491,
+      "step": 500
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 0.17240458107083004,
+      "learning_rate": 1.7295859088069234e-05,
+      "loss": 0.5368,
+      "step": 501
+    },
+    {
+      "epoch": 0.64256,
+      "grad_norm": 0.17840513795401933,
+      "learning_rate": 1.7189501409486062e-05,
+      "loss": 0.5478,
+      "step": 502
+    },
+    {
+      "epoch": 0.64384,
+      "grad_norm": 0.18720267712459818,
+      "learning_rate": 1.7083300154963193e-05,
+      "loss": 0.5616,
+      "step": 503
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 0.16661138773775544,
+      "learning_rate": 1.6977257451436712e-05,
+      "loss": 0.5533,
+      "step": 504
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.181117880126469,
+      "learning_rate": 1.6871375422667375e-05,
+      "loss": 0.5517,
+      "step": 505
+    },
+    {
+      "epoch": 0.64768,
+      "grad_norm": 0.16124311049954437,
+      "learning_rate": 1.6765656189198013e-05,
+      "loss": 0.5399,
+      "step": 506
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 0.1780112598733124,
+      "learning_rate": 1.6660101868311092e-05,
+      "loss": 0.5545,
+      "step": 507
+    },
+    {
+      "epoch": 0.65024,
+      "grad_norm": 0.17321825984782027,
+      "learning_rate": 1.6554714573986324e-05,
+      "loss": 0.5506,
+      "step": 508
+    },
+    {
+      "epoch": 0.65152,
+      "grad_norm": 0.19903568513322475,
+      "learning_rate": 1.6449496416858284e-05,
+      "loss": 0.5463,
+      "step": 509
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.16644287745949193,
+      "learning_rate": 1.6344449504174193e-05,
+      "loss": 0.547,
+      "step": 510
+    },
+    {
+      "epoch": 0.65408,
+      "grad_norm": 0.18250795813297635,
+      "learning_rate": 1.623957593975166e-05,
+      "loss": 0.5406,
+      "step": 511
+    },
+    {
+      "epoch": 0.65536,
+      "grad_norm": 0.1858788673421162,
+      "learning_rate": 1.613487782393661e-05,
+      "loss": 0.5591,
+      "step": 512
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 0.17724801933662296,
+      "learning_rate": 1.6030357253561173e-05,
+      "loss": 0.5418,
+      "step": 513
+    },
+    {
+      "epoch": 0.65792,
+      "grad_norm": 0.1708440658079089,
+      "learning_rate": 1.592601632190169e-05,
+      "loss": 0.5562,
+      "step": 514
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.1784213729096496,
+      "learning_rate": 1.582185711863681e-05,
+      "loss": 0.5394,
+      "step": 515
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 0.180502730233706,
+      "learning_rate": 1.571788172980566e-05,
+      "loss": 0.555,
+      "step": 516
+    },
+    {
+      "epoch": 0.66176,
+      "grad_norm": 0.16436771129623443,
+      "learning_rate": 1.5614092237766006e-05,
+      "loss": 0.5427,
+      "step": 517
+    },
+    {
+      "epoch": 0.66304,
+      "grad_norm": 0.18053318678607747,
+      "learning_rate": 1.5510490721152592e-05,
+      "loss": 0.5492,
+      "step": 518
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 0.18604940199556919,
+      "learning_rate": 1.5407079254835506e-05,
+      "loss": 0.5335,
+      "step": 519
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.19210597805514745,
+      "learning_rate": 1.530385990987863e-05,
+      "loss": 0.5442,
+      "step": 520
+    },
+    {
+      "epoch": 0.66688,
+      "grad_norm": 0.17611283743427303,
+      "learning_rate": 1.5200834753498128e-05,
+      "loss": 0.5322,
+      "step": 521
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 0.22007676229449954,
+      "learning_rate": 1.509800584902108e-05,
+      "loss": 0.5584,
+      "step": 522
+    },
+    {
+      "epoch": 0.66944,
+      "grad_norm": 0.17964636810711826,
+      "learning_rate": 1.499537525584416e-05,
+      "loss": 0.546,
+      "step": 523
+    },
+    {
+      "epoch": 0.67072,
+      "grad_norm": 0.18491688306598894,
+      "learning_rate": 1.489294502939238e-05,
+      "loss": 0.5469,
+      "step": 524
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.20303860785516076,
+      "learning_rate": 1.4790717221077898e-05,
+      "loss": 0.5499,
+      "step": 525
+    },
+    {
+      "epoch": 0.67328,
+      "grad_norm": 0.18536413393916135,
+      "learning_rate": 1.4688693878258991e-05,
+      "loss": 0.5427,
+      "step": 526
+    },
+    {
+      "epoch": 0.67456,
+      "grad_norm": 0.19288854639543657,
+      "learning_rate": 1.4586877044199016e-05,
+      "loss": 0.5576,
+      "step": 527
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 0.19998872597922154,
+      "learning_rate": 1.4485268758025466e-05,
+      "loss": 0.5337,
+      "step": 528
+    },
+    {
+      "epoch": 0.67712,
+      "grad_norm": 0.2356208291248628,
+      "learning_rate": 1.4383871054689213e-05,
+      "loss": 0.5457,
+      "step": 529
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.19146077814425122,
+      "learning_rate": 1.4282685964923642e-05,
+      "loss": 0.549,
+      "step": 530
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 0.23473975201728703,
+      "learning_rate": 1.4181715515204095e-05,
+      "loss": 0.5377,
+      "step": 531
+    },
+    {
+      "epoch": 0.68096,
+      "grad_norm": 0.196745306404929,
+      "learning_rate": 1.4080961727707184e-05,
+      "loss": 0.5458,
+      "step": 532
+    },
+    {
+      "epoch": 0.68224,
+      "grad_norm": 0.1723810980465925,
+      "learning_rate": 1.398042662027035e-05,
+      "loss": 0.5463,
+      "step": 533
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 0.2148768848114498,
+      "learning_rate": 1.3880112206351475e-05,
+      "loss": 0.5511,
+      "step": 534
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.19580634261932736,
+      "learning_rate": 1.3780020494988446e-05,
+      "loss": 0.5458,
+      "step": 535
+    },
+    {
+      "epoch": 0.68608,
+      "grad_norm": 0.18528206536801226,
+      "learning_rate": 1.3680153490759073e-05,
+      "loss": 0.5513,
+      "step": 536
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 0.1833258102759667,
+      "learning_rate": 1.3580513193740818e-05,
+      "loss": 0.5436,
+      "step": 537
+    },
+    {
+      "epoch": 0.68864,
+      "grad_norm": 0.18879719857423252,
+      "learning_rate": 1.3481101599470794e-05,
+      "loss": 0.5381,
+      "step": 538
+    },
+    {
+      "epoch": 0.68992,
+      "grad_norm": 0.18597088087253308,
+      "learning_rate": 1.3381920698905787e-05,
+      "loss": 0.5519,
+      "step": 539
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.19724256330188028,
+      "learning_rate": 1.328297247838241e-05,
+      "loss": 0.5499,
+      "step": 540
+    },
+    {
+      "epoch": 0.69248,
+      "grad_norm": 0.1740013114346114,
+      "learning_rate": 1.3184258919577269e-05,
+      "loss": 0.5413,
+      "step": 541
+    },
+    {
+      "epoch": 0.69376,
+      "grad_norm": 0.18816042265895747,
+      "learning_rate": 1.3085781999467303e-05,
+      "loss": 0.5257,
+      "step": 542
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 0.18327617156382764,
+      "learning_rate": 1.2987543690290222e-05,
+      "loss": 0.5501,
+      "step": 543
+    },
+    {
+      "epoch": 0.69632,
+      "grad_norm": 0.17743894174718156,
+      "learning_rate": 1.2889545959504939e-05,
+      "loss": 0.5423,
+      "step": 544
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.1775283974897721,
+      "learning_rate": 1.2791790769752232e-05,
+      "loss": 0.5393,
+      "step": 545
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 0.16200081974294722,
+      "learning_rate": 1.2694280078815382e-05,
+      "loss": 0.5583,
+      "step": 546
+    },
+    {
+      "epoch": 0.70016,
+      "grad_norm": 0.16722111373810458,
+      "learning_rate": 1.2597015839581033e-05,
+      "loss": 0.5417,
+      "step": 547
+    },
+    {
+      "epoch": 0.70144,
+      "grad_norm": 0.17064698518444713,
+      "learning_rate": 1.2500000000000006e-05,
+      "loss": 0.5416,
+      "step": 548
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 0.18207070316671944,
+      "learning_rate": 1.2403234503048319e-05,
+      "loss": 0.5552,
+      "step": 549
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.17186248314141456,
+      "learning_rate": 1.230672128668831e-05,
+      "loss": 0.5503,
+      "step": 550
+    },
+    {
+      "epoch": 0.70528,
+      "grad_norm": 0.18897462643935614,
+      "learning_rate": 1.2210462283829755e-05,
+      "loss": 0.5523,
+      "step": 551
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 0.18474441376802078,
+      "learning_rate": 1.2114459422291205e-05,
+      "loss": 0.543,
+      "step": 552
+    },
+    {
+      "epoch": 0.70784,
+      "grad_norm": 0.1726434041865777,
+      "learning_rate": 1.2018714624761352e-05,
+      "loss": 0.5342,
+      "step": 553
+    },
+    {
+      "epoch": 0.70912,
+      "grad_norm": 0.17218934161002325,
+      "learning_rate": 1.1923229808760564e-05,
+      "loss": 0.5422,
+      "step": 554
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.18953711573549967,
+      "learning_rate": 1.1828006886602422e-05,
+      "loss": 0.5406,
+      "step": 555
+    },
+    {
+      "epoch": 0.71168,
+      "grad_norm": 0.1880902181302069,
+      "learning_rate": 1.1733047765355466e-05,
+      "loss": 0.5462,
+      "step": 556
+    },
+    {
+      "epoch": 0.71296,
+      "grad_norm": 0.1596287230354233,
+      "learning_rate": 1.1638354346804971e-05,
+      "loss": 0.5388,
+      "step": 557
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 0.17158572664682273,
+      "learning_rate": 1.154392852741491e-05,
+      "loss": 0.5527,
+      "step": 558
+    },
+    {
+      "epoch": 0.71552,
+      "grad_norm": 0.1804729952348338,
+      "learning_rate": 1.1449772198289905e-05,
+      "loss": 0.5426,
+      "step": 559
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.1791816734451783,
+      "learning_rate": 1.1355887245137383e-05,
+      "loss": 0.5588,
+      "step": 560
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 0.16584626596374333,
+      "learning_rate": 1.126227554822985e-05,
+      "loss": 0.5366,
+      "step": 561
+    },
+    {
+      "epoch": 0.71936,
+      "grad_norm": 0.1727030219272147,
+      "learning_rate": 1.116893898236716e-05,
+      "loss": 0.5422,
+      "step": 562
+    },
+    {
+      "epoch": 0.72064,
+      "grad_norm": 0.17521577149409487,
+      "learning_rate": 1.1075879416839023e-05,
+      "loss": 0.5407,
+      "step": 563
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 0.27313054071540727,
+      "learning_rate": 1.0983098715387526e-05,
+      "loss": 0.5456,
+      "step": 564
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.16784636956770904,
+      "learning_rate": 1.089059873616988e-05,
+      "loss": 0.5552,
+      "step": 565
+    },
+    {
+      "epoch": 0.72448,
+      "grad_norm": 0.1733610583523821,
+      "learning_rate": 1.0798381331721109e-05,
+      "loss": 0.5462,
+      "step": 566
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 0.1840396076908537,
+      "learning_rate": 1.0706448348917006e-05,
+      "loss": 0.5358,
+      "step": 567
+    },
+    {
+      "epoch": 0.72704,
+      "grad_norm": 0.1655608413544908,
+      "learning_rate": 1.061480162893716e-05,
+      "loss": 0.5423,
+      "step": 568
+    },
+    {
+      "epoch": 0.72832,
+      "grad_norm": 0.17123662403250298,
+      "learning_rate": 1.052344300722803e-05,
+      "loss": 0.5468,
+      "step": 569
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.17550528875066398,
+      "learning_rate": 1.043237431346622e-05,
+      "loss": 0.5454,
+      "step": 570
+    },
+    {
+      "epoch": 0.73088,
+      "grad_norm": 0.16328379269859897,
+      "learning_rate": 1.0341597371521825e-05,
+      "loss": 0.5514,
+      "step": 571
+    },
+    {
+      "epoch": 0.73216,
+      "grad_norm": 0.16650225308818498,
+      "learning_rate": 1.0251113999421935e-05,
+      "loss": 0.5432,
+      "step": 572
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 0.17350055029708436,
+      "learning_rate": 1.016092600931414e-05,
+      "loss": 0.5409,
+      "step": 573
+    },
+    {
+      "epoch": 0.73472,
+      "grad_norm": 0.1596212141180814,
+      "learning_rate": 1.0071035207430352e-05,
+      "loss": 0.5367,
+      "step": 574
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.2658852675551339,
+      "learning_rate": 9.981443394050525e-06,
+      "loss": 0.5595,
+      "step": 575
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 0.159704372500189,
+      "learning_rate": 9.892152363466691e-06,
+      "loss": 0.5357,
+      "step": 576
+    },
+    {
+      "epoch": 0.73856,
+      "grad_norm": 0.1651078900711754,
+      "learning_rate": 9.803163903946951e-06,
+      "loss": 0.5517,
+      "step": 577
+    },
+    {
+      "epoch": 0.73984,
+      "grad_norm": 0.1787435238987854,
+      "learning_rate": 9.714479797699694e-06,
+      "loss": 0.548,
+      "step": 578
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 0.1652971919910291,
+      "learning_rate": 9.626101820837927e-06,
+      "loss": 0.5539,
+      "step": 579
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.1639694735505629,
+      "learning_rate": 9.538031743343628e-06,
+      "loss": 0.5451,
+      "step": 580
+    },
+    {
+      "epoch": 0.74368,
+      "grad_norm": 0.1647232239012586,
+      "learning_rate": 9.450271329032404e-06,
+      "loss": 0.5505,
+      "step": 581
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 0.16460370416547013,
+      "learning_rate": 9.362822335518063e-06,
+      "loss": 0.553,
+      "step": 582
+    },
+    {
+      "epoch": 0.74624,
+      "grad_norm": 0.1697639854685921,
+      "learning_rate": 9.275686514177507e-06,
+      "loss": 0.5502,
+      "step": 583
+    },
+    {
+      "epoch": 0.74752,
+      "grad_norm": 0.16820172586994564,
+      "learning_rate": 9.18886561011557e-06,
+      "loss": 0.5444,
+      "step": 584
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.16239199386354092,
+      "learning_rate": 9.102361362130132e-06,
+      "loss": 0.5386,
+      "step": 585
+    },
+    {
+      "epoch": 0.75008,
+      "grad_norm": 0.17069749707133253,
+      "learning_rate": 9.01617550267726e-06,
+      "loss": 0.5407,
+      "step": 586
+    },
+    {
+      "epoch": 0.75136,
+      "grad_norm": 0.17402297905951,
+      "learning_rate": 8.930309757836517e-06,
+      "loss": 0.5508,
+      "step": 587
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 0.16830124871078894,
+      "learning_rate": 8.844765847276432e-06,
+      "loss": 0.5258,
+      "step": 588
+    },
+    {
+      "epoch": 0.75392,
+      "grad_norm": 0.15842747930020387,
+      "learning_rate": 8.759545484219984e-06,
+      "loss": 0.5448,
+      "step": 589
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.16546914193605752,
+      "learning_rate": 8.67465037541038e-06,
+      "loss": 0.5377,
+      "step": 590
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 0.1606695724745096,
+      "learning_rate": 8.590082221076765e-06,
+      "loss": 0.5221,
+      "step": 591
+    },
+    {
+      "epoch": 0.75776,
+      "grad_norm": 0.1632549785956385,
+      "learning_rate": 8.505842714900297e-06,
+      "loss": 0.5423,
+      "step": 592
+    },
+    {
+      "epoch": 0.75904,
+      "grad_norm": 0.16168391888394712,
+      "learning_rate": 8.421933543980126e-06,
+      "loss": 0.5458,
+      "step": 593
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 0.1756546275289762,
+      "learning_rate": 8.338356388799637e-06,
+      "loss": 0.5442,
+      "step": 594
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 0.16963428171763828,
+      "learning_rate": 8.255112923192835e-06,
+      "loss": 0.5459,
+      "step": 595
+    },
+    {
+      "epoch": 0.76288,
+      "grad_norm": 0.15718915668817968,
+      "learning_rate": 8.172204814310742e-06,
+      "loss": 0.5449,
+      "step": 596
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 0.16536550501030137,
+      "learning_rate": 8.089633722588103e-06,
+      "loss": 0.5387,
+      "step": 597
+    },
+    {
+      "epoch": 0.76544,
+      "grad_norm": 0.18318586900512912,
+      "learning_rate": 8.007401301710022e-06,
+      "loss": 0.5512,
+      "step": 598
+    },
+    {
+      "epoch": 0.76672,
+      "grad_norm": 0.16502774685748536,
+      "learning_rate": 7.92550919857896e-06,
+      "loss": 0.5369,
+      "step": 599
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.15963079625942453,
+      "learning_rate": 7.843959053281663e-06,
+      "loss": 0.5445,
+      "step": 600
+    },
+    {
+      "epoch": 0.76928,
+      "grad_norm": 0.18142281673636804,
+      "learning_rate": 7.762752499056358e-06,
+      "loss": 0.5521,
+      "step": 601
+    },
+    {
+      "epoch": 0.77056,
+      "grad_norm": 0.17003382425447292,
+      "learning_rate": 7.681891162260015e-06,
+      "loss": 0.5481,
+      "step": 602
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 0.15682351369009967,
+      "learning_rate": 7.60137666233583e-06,
+      "loss": 0.5427,
+      "step": 603
+    },
+    {
+      "epoch": 0.77312,
+      "grad_norm": 0.16900483827947543,
+      "learning_rate": 7.521210611780716e-06,
+      "loss": 0.5491,
+      "step": 604
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 0.16933634809215875,
+      "learning_rate": 7.441394616113062e-06,
+      "loss": 0.5321,
+      "step": 605
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 0.15523375999123107,
+      "learning_rate": 7.361930273840581e-06,
+      "loss": 0.5359,
+      "step": 606
+    },
+    {
+      "epoch": 0.77696,
+      "grad_norm": 0.16476118124777908,
+      "learning_rate": 7.2828191764282515e-06,
+      "loss": 0.5405,
+      "step": 607
+    },
+    {
+      "epoch": 0.77824,
+      "grad_norm": 0.15880451278636304,
+      "learning_rate": 7.20406290826649e-06,
+      "loss": 0.5341,
+      "step": 608
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 0.16601175165765483,
+      "learning_rate": 7.125663046639386e-06,
+      "loss": 0.5354,
+      "step": 609
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.1594428201948071,
+      "learning_rate": 7.047621161693152e-06,
+      "loss": 0.5414,
+      "step": 610
+    },
+    {
+      "epoch": 0.78208,
+      "grad_norm": 0.15903842435489843,
+      "learning_rate": 6.969938816404639e-06,
+      "loss": 0.537,
+      "step": 611
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 0.16373860391653333,
+      "learning_rate": 6.892617566550044e-06,
+      "loss": 0.5324,
+      "step": 612
+    },
+    {
+      "epoch": 0.78464,
+      "grad_norm": 0.14806851457758538,
+      "learning_rate": 6.815658960673782e-06,
+      "loss": 0.5423,
+      "step": 613
+    },
+    {
+      "epoch": 0.78592,
+      "grad_norm": 0.16162392173499782,
+      "learning_rate": 6.739064540057424e-06,
+      "loss": 0.5379,
+      "step": 614
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.1556598212425785,
+      "learning_rate": 6.662835838688864e-06,
+      "loss": 0.5409,
+      "step": 615
+    },
+    {
+      "epoch": 0.78848,
+      "grad_norm": 0.1545780708161842,
+      "learning_rate": 6.586974383231573e-06,
+      "loss": 0.5371,
+      "step": 616
+    },
+    {
+      "epoch": 0.78976,
+      "grad_norm": 0.15481486085975593,
+      "learning_rate": 6.511481692994076e-06,
+      "loss": 0.5514,
+      "step": 617
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 0.15187060933009017,
+      "learning_rate": 6.436359279899426e-06,
+      "loss": 0.5352,
+      "step": 618
+    },
+    {
+      "epoch": 0.79232,
+      "grad_norm": 0.1627175375732009,
+      "learning_rate": 6.361608648455039e-06,
+      "loss": 0.54,
+      "step": 619
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.1550038050293079,
+      "learning_rate": 6.28723129572247e-06,
+      "loss": 0.5436,
+      "step": 620
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 0.1605066763261186,
+      "learning_rate": 6.2132287112875e-06,
+      "loss": 0.5401,
+      "step": 621
+    },
+    {
+      "epoch": 0.79616,
+      "grad_norm": 0.15630165231383042,
+      "learning_rate": 6.1396023772302465e-06,
+      "loss": 0.5353,
+      "step": 622
+    },
+    {
+      "epoch": 0.79744,
+      "grad_norm": 0.14818753466561113,
+      "learning_rate": 6.066353768095504e-06,
+      "loss": 0.551,
+      "step": 623
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 0.17322071941695266,
+      "learning_rate": 5.993484350863246e-06,
+      "loss": 0.5372,
+      "step": 624
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17299311811164592,
+      "learning_rate": 5.92099558491917e-06,
+      "loss": 0.5491,
+      "step": 625
+    },
+    {
+      "epoch": 0.80128,
+      "grad_norm": 0.15432256972090858,
+      "learning_rate": 5.848888922025553e-06,
+      "loss": 0.5451,
+      "step": 626
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 0.1796991664902234,
+      "learning_rate": 5.777165806292109e-06,
+      "loss": 0.542,
+      "step": 627
+    },
+    {
+      "epoch": 0.80384,
+      "grad_norm": 0.15716216593694202,
+      "learning_rate": 5.7058276741471236e-06,
+      "loss": 0.5359,
+      "step": 628
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 781,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 157,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 780227321331712.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}