diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.982278481012658,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 7.215950835338928,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.1957,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.951795652079414,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.2203,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 7.523960371689583,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.218,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.853762257148823,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.1747,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.634178028008215,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.158,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 6.581350028552881,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.1936,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 5.401066262949182,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.1581,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.822083989492024,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.1415,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.801975645647603,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.0949,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.8113957981481255,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.0992,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.396896207421952,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.0766,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.626306928379588,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.0524,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 3.979470952810443,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.0625,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.5239177878677994,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.0057,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.291941205094637,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.0305,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.254875987483683,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.9935,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.450745971873278,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.9412,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.353809865265511,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.9543,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.205701894569386,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.9116,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.8787284276474054,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.9309,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.4800529786531944,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.9272,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.1872884915755955,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.9289,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.4609183333885172,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.8969,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.3788348172578622,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.8715,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.391587977385867,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.8869,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.1241090250093615,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.8856,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.8976934169939406,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.8568,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.0815374932478194,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.8562,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.1584419688312468,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8485,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8080957470450341,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.847,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.8334264218605529,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.8292,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.883407814328604,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.8316,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.8822073493969109,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.8201,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8203165192751278,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.8455,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.7543647967513224,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.8067,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.839302133167973,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.8383,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.8434913390382298,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.8366,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.8450926187888216,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.8015,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.7488158827575802,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.82,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.7630023916721811,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.812,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.8204193780158495,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8101,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.7506365662488276,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.7845,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.7330085934695888,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.7888,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.763164308099687,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.8131,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.6784221434800476,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.8132,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.7698407632823899,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.7907,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7014688642819048,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.794,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.7625428088060089,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.7949,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.740892683915829,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.777,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.7529510705031943,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.7786,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.8062185298209003,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.8115,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.9737877542862647,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.7928,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.8426313892089375,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.7819,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.8624032841992368,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.7731,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.8062058533931598,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.7991,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.8464754431469301,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.8024,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.7540489611264224,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.7712,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.7207169263602223,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.7476,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7379426113031257,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7815,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.8386650796953586,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.832,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.7843745157556686,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.8241,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.7285803944561876,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.7902,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.7210211925976612,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.7702,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.7337356756622401,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.7841,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.77189375980598,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.762,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 1.0413606713138686,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.7888,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.7580256521398194,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.7635,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 1.1900923184061862,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.7967,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.0107478479466288,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.791,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.8986142022269631,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.7597,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.1452509920226213,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.7767,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.9007139216097757,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.7946,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.3812767587261026,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.792,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.045744838864834,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.7705,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.4437292278173377,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.7748,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.399469753542205,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.7937,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.1093230648632673,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.7503,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.0388684025408355,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.7525,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2555991443537202,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.7795,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.937120990052005,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7942,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.1520935964191388,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.8143,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.0351317997448832,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7836,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.0070945454832643,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.7739,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.9472526330880872,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.7675,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.8391276635911974,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.7444,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.9007357290353687,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.7765,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.8176887710315593,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.7503,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.8993774457575491,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.7793,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.9137949236484697,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.754,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.9191714112953735,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.7336,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.8723887858888314,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.7711,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.375255764790935,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.7924,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.0631423115215777,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.7784,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.0206160164641667,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7639,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.110378896278531,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.7748,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.9806575856484651,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.7775,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.9354261120792736,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7566,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.7429270313641705,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.7544,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.9455007510316968,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.7826,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.1245784400779093,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.7706,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.196743892533454,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.7848,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.786511727488043,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.7584,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.8479583124188059,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7553,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.7963499116263102,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.7464,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.0504012049754037,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.7613,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.823501012592923,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.7618,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.8395573388359885,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.758,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.0248898262854351,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.7433,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.015565801574212,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.747,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.8055583103817744,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.781,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.1307191915071293,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.7556,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.8899692072590482,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.7697,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.8210675296686604,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.7399,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.1858910352515222,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.759,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.8650709834265636,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.7631,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.9924397181152211,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.7844,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.2513719075894412,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.8028,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.8258672403771011,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.7844,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.1152417014139089,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7698,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.1547100306905655,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.7563,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.9826256328136171,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7547,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.182971420114039,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.7617,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.9673562392050694,
+      "learning_rate": 4e-05,
+      "loss": 0.7385,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.9118963560322927,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.7686,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.0857396601917215,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7619,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.7969535657262729,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7458,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.9622097080531534,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.7811,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.7610465865582012,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.7437,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.8098176703957123,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7814,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.8232545215636232,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.7525,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.840336025547633,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.7487,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.8227155053617193,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.753,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.0964912311528128,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.7341,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.9246514177286603,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.7822,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.0557768763064626,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7696,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.1630099049720832,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.7761,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.8360771210570436,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.7648,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.1419877331006703,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.727,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.765059104453185,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.7693,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.1868126978331486,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7992,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.7398860754439288,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.7204,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.9069785380803419,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.7882,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.8429477133357086,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.7465,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.0642799344555212,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.755,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.7538323051821632,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7797,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.8503215047121716,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.7528,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.7306154695881446,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.7796,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.9717068620276939,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.7893,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.8139879068886657,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.7578,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.0854353377398653,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.7691,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.8064742638461486,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.7543,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.7229915655521082,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.7375,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.9800572508653468,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.7733,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.6168629105554387,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.7366,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.0061219308477918,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.742,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.7142350028224932,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.7488,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.8444024055710673,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.7326,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8352707223791813,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.7373,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.9341326483080068,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.7834,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.8310598917986102,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.7451,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.0627156317647306,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.7681,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.762950052368729,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.7916,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.8160431231981005,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.749,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.7799763777405676,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.7859,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.7500838539059973,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.7656,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.8383758927312486,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.774,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.6988404266411504,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.781,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.8042683477400006,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.7185,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.8659024523127444,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.7468,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.8148549641649042,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.7434,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.8343797382932288,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.7649,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.7680049046244791,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.7578,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.7639433425932247,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7794,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.7340301884745243,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.7768,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.9259421072605722,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.7508,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.7716823763261841,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.7376,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.83771479013137,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.7473,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.6803803883946647,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.7439,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.8329594806531534,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.7921,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.8121414756457791,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.7725,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.6765691026722147,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.7568,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.7500579852037887,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.7415,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.6233033606602053,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.7268,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.8160862986865002,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.7863,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.6690167095220626,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.7691,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.9175563981802893,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.7215,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.675897594020786,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.7514,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.7155538926306435,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.7782,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.7948919522191639,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.7405,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.6652592847804493,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.7531,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.7659575514157849,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.7511,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.6213680843823318,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.723,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.6694013944821202,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.7464,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.7116200514682692,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.7334,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.6270680215925424,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.7586,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.7455664685354269,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.7612,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.7180571822771513,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.7699,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.7573097767279965,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.7433,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.7827723862537322,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.7638,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.8540451424411167,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.7375,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.8249371348734224,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.7577,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.6936094835268473,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.7359,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.857811278004629,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.7455,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.828011276699999,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.7764,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.7491043097627976,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7503,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 1.1678031343913937,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.7563,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.7059094948910418,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.7725,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.9144199515841348,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.764,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.7132249679544752,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.7731,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.786241963894141,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.7431,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.7584698951811515,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.737,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.7845960046507491,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.7607,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.8025466340209193,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.7387,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.7588065709881274,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.7522,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.8067842562891061,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.7764,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.6796695256373263,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.7514,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.7719739366400243,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.7294,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.7991764857089436,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.7522,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.6786486683731796,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.7777,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.902911681496041,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.7534,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.695915769917628,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.7314,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.8569666763761161,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.6872,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.6921110855856338,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.7426,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.9514187757815081,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.7472,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.8288934411719919,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.7526,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.7759474805120715,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.7109,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.7305474997747919,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.7679,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.891438276190769,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.7519,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.746068142865399,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.7686,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.8759732151990794,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.7412,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.7415076340601627,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.7667,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.8556977672938606,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.7401,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.766338683751792,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.7345,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.6692235675385891,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.7705,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.8157546011218765,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.7286,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.7221237529377361,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.7508,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7035611176950511,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.7489,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.7955037250306818,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.7438,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.6932628916257914,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.7606,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.8987091838502814,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.7455,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.721323087934312,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.7514,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.8387375169239146,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.7127,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.7599701137371306,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.7465,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.7525679814769082,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.7375,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.8962527308862397,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.7621,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.7995357619832281,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.7383,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.8408584881295867,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.8351,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 1.0092410598120183,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.5772,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.6859890076575119,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.548,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.9721602496556997,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5357,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.8586725495081553,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5508,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.7124550369368671,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.534,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.8905423792723123,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5676,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.7327164502998588,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5443,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.9471396344162789,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5733,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.7050623169840646,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5208,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.9015870496577499,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5413,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.7642313558047085,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5509,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.9603542842118604,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5821,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.7526484156773239,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5454,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.7515089465445806,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5606,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.7183270608985367,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.5335,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.7311122024214733,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5398,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.677033741818442,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5484,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.7345811542653122,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5564,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.698102837670455,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.535,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.7449034699718642,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5434,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.6858726471048094,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5208,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.7054332445146511,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5344,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.7270258332095847,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5373,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.7428920198573847,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5235,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.6671403224898788,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5701,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.7163715054790031,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5154,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.7052320281635165,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5607,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.6935490684560304,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5266,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.7086095162356304,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5193,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.6086744612659778,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5095,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.7324416795708132,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5668,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.6521611543980032,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5336,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.6917138095024545,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5282,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.606106389435255,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5454,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.6346348810144748,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5487,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.8024019193199146,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5504,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.6018286684501064,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5286,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.8297515842096755,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.567,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.7254969954012782,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5419,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.6624833404346386,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5446,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.7109376129387933,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5524,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.662247087355836,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.544,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.6715028555500593,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5077,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.7335516520545975,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5324,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.885630417717264,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.6075,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.7446988960708469,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5387,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.6495593663564753,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5563,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.6641823236289368,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5189,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.6887616219574604,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5415,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.7464691321022285,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5493,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.7215320464212671,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5499,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.6686599582789137,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5381,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.6828892926629933,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5543,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.7072612520626365,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5423,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.6447135174621457,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5515,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.7089217291930017,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5427,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.7394892652596364,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5696,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.5983107446354792,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5437,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.7509069600344553,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5425,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.5931728658389552,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5542,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.9226249216012835,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5515,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.7132111304572716,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5377,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.6704199537535175,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.5397,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.5824912208467539,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.5441,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.68675062012765,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.525,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.6843841877593598,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5648,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.9060640651170169,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5819,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.5944833775630098,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5176,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8793417512229861,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5696,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.6372737026775087,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5467,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.7690116766146395,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5292,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.6937358124436743,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.557,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.7389063694016071,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5698,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.6695018185339149,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5383,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.6374626659820671,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.5574,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.7133168093118089,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5225,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.5992148903876257,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5444,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.7079535831253218,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5244,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.802054748425032,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5588,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.7126830075191588,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5679,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.8382656920407481,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5118,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.6988126266771416,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5757,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.9002185749010725,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5502,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.6490752546474386,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5371,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.7923061718413087,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.557,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.611280400063123,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5326,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.7364505387176472,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5417,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.7108354409830835,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5385,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.5959772493970517,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5359,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.8014465848210842,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5811,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.6134180854483254,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5457,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.6966864507659891,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5585,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.6357378871590282,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.5151,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.6829328078328969,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5694,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.6364467363768119,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5649,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.6485611201567729,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5418,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.6249064626657965,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.5597,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.7119845525615972,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.569,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.5906698834768196,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5534,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.6362173337316753,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5591,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.6366094010152824,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5839,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.5965631987822544,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5381,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.6191906666094317,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5241,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.6743460292740124,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5686,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.5990800462646027,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.556,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.631760986424611,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5486,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.6093209421562892,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5693,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.6640165817743495,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5647,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.6338204410703887,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5436,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.669622503544241,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5532,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.5999451802419344,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5376,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.7329341087128348,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.6102,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.6744673319485851,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5292,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.6538944503869045,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5671,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.6604789432091305,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5411,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.7827680290020442,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5641,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.6675181040096452,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5874,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.7745659708668124,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5185,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.6817031073872869,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5871,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.6468096036962667,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5749,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.6912643421719086,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5568,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.6437197337774866,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5355,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.6725589140527675,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5588,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.7249456349638312,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5813,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.6593251020022628,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5424,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.7341106379243226,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5493,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.6988279874971601,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5782,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.6441670358706216,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5543,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.6102194327413953,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5508,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.7337829344648867,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.5382,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.5776743401164992,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.559,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.7269906579807949,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5421,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.6276680705827339,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5671,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.6785363957356492,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.5877,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.7068335676404874,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5564,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.5786049083250571,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5262,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.72644102741444,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5634,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.640860297030729,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.5462,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.575164510073355,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5483,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.6163266004118536,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.5474,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.634510915816311,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5601,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.5831483401791605,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5357,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.6060703457573784,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5562,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.6541507443858636,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5367,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.6001584199368653,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5295,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.6238940839014454,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.561,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.6888982195686422,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5848,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6225997216011292,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5261,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.6012744764969128,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5594,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.6342766912361806,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.5802,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.686532620703637,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5604,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.6357582931199729,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.535,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.6488302766294088,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5539,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.5954061126920128,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5191,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.6503790770894794,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5861,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.6207011578140214,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5409,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.6141851831633796,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5294,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.6837268665117877,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.5578,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.6086490732036299,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5409,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.5907890895236158,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5694,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.5918583159337685,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5738,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.6441508829486058,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.5262,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.6063978084143573,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5449,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.6514082256287121,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5468,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.7680005876060783,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5808,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.5563075259387089,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.497,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.7846254825072837,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5569,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.6312008658956921,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5618,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.6489379676485724,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5687,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.6461182669757004,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.5538,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.591907499890267,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.5841,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.6237303331619759,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5606,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.6318364327874199,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5557,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.6750254659429133,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5543,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.6411715228398577,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5446,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.639537908547554,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5465,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.6362719190654301,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5401,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.6610784491793237,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.5684,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.606543003498884,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5626,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.7348423074691008,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.5517,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.5928451583334002,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5199,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.6343169253124734,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.5533,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.656916799580976,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5472,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.6910872855720994,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5623,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.6623815845939316,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5449,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.7799234752562278,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5213,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.6898568262729693,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5708,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.7311270114791362,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.5469,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.6596256904950155,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5485,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.6301631145191625,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5679,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.6548219739100598,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5863,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.6893687340882793,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5478,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.6565583253974497,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5751,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.7714048583982069,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5461,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.663140843182443,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5446,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.6605785595281253,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5525,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.6209887935305272,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5249,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.7280701912178379,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.5821,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.6012703847483686,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.557,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.5737865319467917,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5456,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.6722341990961275,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.5745,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.5917109617548341,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5303,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.6118172253994537,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5513,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.6566872048910005,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5435,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.6615010772765102,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5867,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.6826061435527357,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5301,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.631209865867508,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.5968,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.7377643037738634,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.549,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.6461977041300357,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5716,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.6676564775610883,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5915,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.6899920872082783,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5644,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.6601551731082249,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5178,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.7614526515434983,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5672,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.7403162522286417,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5592,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.7095832826619407,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.531,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.7320976128736354,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5682,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.7469765171469339,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5728,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.769120807877647,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5751,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.786710513447379,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5681,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.9645969712398451,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5797,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.5933013768005059,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5652,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.9274595017744763,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.5535,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6243429256355202,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.5236,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.6727476813517504,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5465,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.6655607943620454,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.5644,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.6280625880196307,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5314,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6011500594491552,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5416,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.701423236212655,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5676,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.5867144197321308,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5454,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.6853280372459285,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5402,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.6575192083417176,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5644,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.551425496912083,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5269,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.7164098763225356,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.573,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.585650154557475,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.541,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.6326551421912959,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5551,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.6715289620821129,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.582,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.597288718706357,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5506,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.5899009840220579,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5441,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.6003532745432277,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5675,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.5686872194830235,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5352,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.6962335080237658,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5445,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.6542656482550815,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.5829,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.6263850172077654,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5556,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.7826636752521474,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5509,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.5576183036102228,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5629,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.6727525527770221,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5521,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.8713806337128074,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.5948,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.768474614045476,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3122,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.724439768163998,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.3267,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 1.0609572411612018,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3346,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.6515118934567621,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.299,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.7294777209219776,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3091,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.6670756734937525,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3125,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.5910491868419896,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.2994,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.6560432143048389,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3049,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.6382181059833583,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.3181,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.6598629029745922,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3229,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.602431205890049,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3173,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.625447325307316,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.355,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.6365362632369358,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.2945,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.5605598685503446,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.2789,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.6196565493813223,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3175,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.5895099645835674,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3018,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.6230360354323756,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.3149,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.5627347357830011,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.3088,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.5798944752815284,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3191,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.5914939690176643,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.2947,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.5739563170598994,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3127,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.5578883802997592,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.2987,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.5792269430539614,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3313,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.616300522150984,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3143,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.5721528800311082,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3178,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.5447502041848296,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3017,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.5538410248299674,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.2923,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.6496600229723855,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.2948,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.543251142513437,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3093,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.7230515933528445,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.303,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.5917539802052064,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3089,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.6608579196539608,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3179,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.6212385754778622,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3047,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.5331301352616202,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3039,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.6080487178597063,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3219,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.5709845033407467,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3167,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.5333310618988124,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.2938,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.5727196392137356,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.3255,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.5617701108197453,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3257,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.5286625907208484,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.2858,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.5248305985056525,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3055,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.491385669362493,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3106,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.5538410443458033,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3134,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.5322989300192444,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.313,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.5718269780795917,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.3222,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.5553291609347274,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3208,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.5513712268668706,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.3185,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.5777713995158046,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3164,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.546275364530869,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.2979,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.5561406977706093,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3178,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.5083798632841602,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3035,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.6211704670390553,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3272,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.517958948647773,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.3068,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.5729466982693455,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3197,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.5618082928478205,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3015,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.5661897008239426,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3292,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.510715156048581,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.2992,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.6370198201478348,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3292,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5037617737275076,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.2824,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.5936171599920654,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3554,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.5406948414542817,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3275,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.511744885070637,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.2783,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.536099642965874,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3168,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.5029653276607926,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3019,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.5884757466213062,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.34,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.520255933409593,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3251,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5407223703151908,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.3182,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.530629416592553,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3315,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.5203172123757867,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.3111,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.5729107434470154,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.325,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.5287091017665658,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3121,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.5472550129619727,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3247,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.5120526961442767,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3094,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.5520953791341723,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3072,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.5163313004002796,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.2989,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.5378701628816427,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3141,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.5678100686529489,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3395,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.5067521500855157,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.3128,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.5428275410355615,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3199,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.5687357509302027,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3236,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5309706344344868,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3068,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.5475889974976967,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3114,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.6225001994189339,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3257,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.5150391722079557,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3037,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.5542718254976255,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.306,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.5319083517610066,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.315,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.5269376879591168,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3157,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.5443674695461532,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3061,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.51936920702072,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.3039,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.5496336424306127,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3166,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.57044670459874,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3081,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.552161963457199,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3271,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.5451029808128596,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.3081,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.552379628527005,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.3169,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.547058640798325,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3042,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.5514042983882244,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3178,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.5482061493557103,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3416,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.4889326668709904,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.3005,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.580686992953589,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3185,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.5454466947864177,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3301,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.5208148545698599,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.2969,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.5476100169053293,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3362,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.5908825104237063,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3305,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.478873908957015,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.2964,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.5656676247353748,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3134,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.5558405919119426,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3363,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.5387417212224908,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3229,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.5522149831333671,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3149,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.5668679738845236,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3416,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.5101204300207784,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3128,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.5429195532071756,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3238,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.5584875073535334,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.3196,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5099760916804554,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3089,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.5382860449486441,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3166,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.5224463436933465,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3221,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.5819359732982177,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3263,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.4974825526881791,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.287,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.5857687801574174,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.344,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.4755241532579914,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.2904,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5247702847079315,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3023,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.5313345931822261,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3379,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.5157924662434655,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3058,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.5665749057794423,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3285,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5208547054677296,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3144,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.5071856626310227,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3085,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.5326739116059203,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3125,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.5396020918570731,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3565,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.5219529620837907,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3174,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.5776357750382086,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.3197,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.5295874194957199,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3112,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.491698064758808,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.2982,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.5588924116153491,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3063,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.5340686556002208,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3166,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.5683258173191007,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3409,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.5505423378424924,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.331,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.5077059959119629,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3122,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.4847497433171961,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3116,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.48742015969605595,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.2787,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.533456786771811,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3163,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.5874856756325801,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.3509,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.5446326670636257,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3309,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.4824868966462237,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.2869,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.5887303736768245,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.3307,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.609405559066818,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3343,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.5848313357236732,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.3411,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.5497576010117476,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3181,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.5557096080054361,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3015,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.5192799842098724,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3071,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.5442742084388141,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3134,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.5284492689224606,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3041,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.5273362530994866,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.312,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.5426014484530512,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3276,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.49021340186341367,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.2898,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.5445911591729876,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3297,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.5507181829548055,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3272,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.5333953846372035,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.2963,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.550946108346463,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3052,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.5637989858789046,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3464,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.5784937138392847,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3302,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.5787600087600709,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3198,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.4971095693246327,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3042,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.5723272907481335,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3176,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.5349223336637862,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3335,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.5598152745125253,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.2998,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.5529961956694575,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3191,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.5196827082966287,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3275,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.5412454332179051,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3145,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.5224223037017144,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.2902,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.5653226546232407,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3448,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.5446920540562872,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3148,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.571138856047431,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.3027,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.5778488767379067,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3436,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.5709180910453419,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3109,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.5399781259500193,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3032,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.5136369643439518,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3079,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.5753136926463664,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.359,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.5279199898921868,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3218,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.559364962962875,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3397,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.5191397997042324,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.2847,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.5331189404436303,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3116,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.5599903509431613,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.324,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.5299847713413045,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3113,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.5445171685911647,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3331,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.5918349342845937,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3373,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.5012874673698235,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.2978,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.5246220060285013,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3065,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.5128693894425614,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.306,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.5282675879941071,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3183,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.5933588565802755,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.3514,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.5434098000064713,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.3237,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.5478296332975087,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.2908,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.5237510883424333,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3062,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.5600265102921625,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3264,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.6112589531989928,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3163,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.5582361158228212,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3362,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.5479461614901044,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.2998,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.5283805336591955,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3196,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.5489030314716928,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3276,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.5318835408929562,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3088,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.505352717421079,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3129,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.5173743982199239,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3095,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.48852634542399903,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.2868,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.5301147547536279,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.349,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.5318959380424302,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3225,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.5231958034790184,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.3033,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.5566248532494651,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3073,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.5211576544517589,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.3286,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.5200502777863479,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3247,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.509269874181525,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3138,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.510515459437771,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.2963,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.49815691424237946,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3172,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.5051927187524577,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.2957,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.5397556367376644,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.3155,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.5301679733872092,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3286,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.5247194382162513,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.31,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.5243856019330353,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3105,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5111087018852479,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.338,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.5164122449063271,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.2971,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.5579716792473983,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.3236,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.524714061703642,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.3241,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.49763688003463363,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3058,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.5482805732192179,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3186,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.5491111384009572,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.3314,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.4986547788551824,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3013,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.6020150093839174,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.322,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.48055302566308694,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.296,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.5443378921585358,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.33,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.5712827405683075,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3115,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.5334745179771978,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.32,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.5358063317126253,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3155,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.5343190233586678,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.3118,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.5411758406500047,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3313,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.49613621283474085,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.2879,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.5216444844760803,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3203,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.5525932611707103,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3221,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.5323972976912034,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3079,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.5184684605028138,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3014,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.564976014942134,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3239,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.5277276859809008,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3117,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.4946252542881344,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.2895,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.52267754056063,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3272,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.518898853385902,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3012,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.5378770913296064,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.322,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.5085264563454951,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3178,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.4933503072747123,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3079,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.5508606610880606,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.345,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.5708505574561843,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.2792,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.6622246817609273,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.1678,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.527818667174338,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.155,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.45617828115638964,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.1532,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.5497909337556098,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1486,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.7416900607589287,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.1594,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.5912267262544678,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1519,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.5217142679245522,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.1495,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.4399361853610566,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.1392,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.5178513906834047,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1638,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.48248503859059666,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1402,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.4428492865358769,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1349,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.45738039343502895,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1422,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.4621893583522543,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.1542,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.4550183956883198,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1414,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.49128710053486485,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.1541,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.4545103402724971,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1374,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.47868918420508516,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.14,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.4648988601440394,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1414,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.48143955654814724,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.1589,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.4386165444158033,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1463,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.4643101707282186,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1578,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.41172086113711626,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.134,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.4438233952357061,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1595,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.4290550203373105,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1449,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.4221479125816219,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1381,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.4445415485736503,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1478,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.4468772430517728,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.14,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.45774297169839356,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1423,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.493785615588431,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1418,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.470624753484527,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1498,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.46340569706646023,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1419,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.45040108258837747,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.1425,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.44122751205741795,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1375,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.4502739866560015,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1446,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.4416468033220235,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1383,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.4664825933273241,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1594,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.44731395657762113,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1511,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.4388021099500862,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1508,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.4268264711494543,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1374,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.4122557100766812,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.1283,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.44028252824965836,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1418,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.4527757442952481,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1409,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.42548977394589566,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1399,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.43362584039034335,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.141,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.441993793591824,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1404,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.45983804902766795,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.1456,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.4136528975823971,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1428,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.45957449525202193,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1515,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.4324100521245061,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1354,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.4750116730131801,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.1498,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.4413855876922569,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.126,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.43127704111633475,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.1394,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.43555547875001965,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1418,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.4510914893566925,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.1541,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.4283239304486354,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.1452,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.4421674524124903,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1454,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.40668782201026943,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.1278,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.4295480124269499,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1356,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.4439972673079792,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.154,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.4038819367395736,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.1282,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.4992206277400803,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1607,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.4121228316411922,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.1332,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.44070863933570986,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.1373,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.49098759809941156,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1497,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.4758711298604978,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1467,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.46223692436788844,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1545,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.429519357168253,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.145,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.43472781976736674,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1466,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.42652059524546543,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1339,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.42026643307218214,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1339,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.4542239201431375,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1473,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.41787109549434204,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.1201,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.4738259469028753,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.1575,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.42339163307875183,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1434,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.4484060899959026,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1469,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.44211346941800317,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.1487,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.44157587098932266,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.138,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.4394335993490547,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1438,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.440354709243475,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1427,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.42069537777270755,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.1353,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.46712468340819263,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1534,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.43309108003607083,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1397,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.43057235641035724,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1432,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.4340327025130414,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1422,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.41345274468775417,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.1351,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.45804852238035093,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1526,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.40487173144589295,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.1245,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.417390302882966,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.1349,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.4302837318074622,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.1512,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.4285396006243638,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.133,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.4544325908071277,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.1583,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.41128004979292254,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1349,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.4418549854797801,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1567,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.4306056593284109,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.1467,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.39138350725055165,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1329,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.44765842453257004,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.1616,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.44543652003362233,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1514,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.42694349427927714,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1323,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.39952813766323964,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1246,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.44647392197577646,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1445,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.4343154746099269,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1379,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.408822537984061,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1298,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.601185463425,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1803,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.4245720103911882,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1346,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.42320993727494416,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1412,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.4360576809330497,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1454,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.4148700940687818,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.1422,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.4066534796904052,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1296,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.4461979045902081,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.1434,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.4122237318543604,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1347,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.4435740760000773,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.152,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.4300342474719172,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1378,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.4314657266524441,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1258,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.4134116204033093,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1439,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.44444573310995983,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.1523,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.39364697652694186,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1253,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.4587456284638789,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1741,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.4234304833883068,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1412,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.4324395192345209,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.139,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.42210426221739666,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1403,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.4506538542774965,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1629,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.4172695954587673,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1355,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.4588304816228176,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.1444,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.41543156545280663,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1303,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.42936614710955984,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.13,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.46370811990768684,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.154,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.4270412545455666,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1408,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.40858280977878625,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1323,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.4316400394296944,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1319,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.4223127242437643,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1413,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.43654327411320004,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1552,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.4047726282424759,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1321,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.43649550428902917,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1372,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.43609879138194835,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1477,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.42078059549639085,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.14,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.43185939190330513,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1385,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.4378707307356163,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1443,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.4219878501430419,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.1328,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.3972980888030331,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1251,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.4254129078099623,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1407,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.4169992165086016,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1429,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.43849965471948466,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.1474,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.3989157940124167,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.1297,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.41100203981675865,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1338,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.4388368555198181,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1473,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.40331567296630616,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1263,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.45308529501655875,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.1506,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.3972185027299793,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.13,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.407634943489806,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1325,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.44313595165614567,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1567,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.41771185002085964,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1372,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.4392498532001948,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1434,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.41503671033463646,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1302,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.4284978414595947,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1527,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.41035479350815574,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1306,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.420431504690146,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.139,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.4245711280080871,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1431,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.4118575476620266,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.1219,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.45180784959419834,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.15,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.38503898894783295,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1324,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.4244202470733448,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1422,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.42830910577033937,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.1361,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.45374717087111854,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1475,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.41268443487820433,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1462,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.42954235606727875,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1523,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.40800097318454615,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1294,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.39719601875883315,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1271,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.40473327960969885,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1276,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.42170685561090965,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1442,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.43143639116686683,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1419,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.44508160386813805,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1446,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.41005843783559054,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1391,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.4392469724008932,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1366,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.4253406087759566,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1296,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.4521107163646157,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.142,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.40729498387217994,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1309,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.4020955398154602,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1249,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.4107867008535436,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.1323,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.4076630196458353,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.133,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.41132101165555773,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1294,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.4378000442397357,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1392,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.4125949437283563,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1403,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.43455364804738794,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1437,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.41524101376323247,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1329,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.4554702469959059,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1649,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.3986426852463736,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1265,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.42234432297431856,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1413,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.4352294196639608,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1396,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.42819370598040885,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.126,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.43697409709580814,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1469,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.40435107777923296,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1309,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.43114692449387226,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1387,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.4083731829570775,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1233,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.4329682275483319,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1439,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.4169593236877785,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.127,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.4095936215590702,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1305,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.4351908639670904,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1455,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.4266462165875928,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.143,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.39843467927083526,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.1243,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.4135908257346817,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.134,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.4253515096160921,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1478,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.41183484615040916,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1235,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.4298161853166184,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.1442,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.43845341300056784,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1554,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.40945245981873546,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1294,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.4111856097025742,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1393,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.4183208540140859,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1389,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.4117565763431292,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1267,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.41549630062964854,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1447,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.3977673636897959,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1339,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.4095254868915055,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.124,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.43227390531283827,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1462,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.43067890720717483,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.161,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.38265544779208077,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1187,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.4253838791848795,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1527,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.4176867427192324,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1331,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.40320903195521435,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1371,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.4119943436858188,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.133,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.4126023242319009,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.1349,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.40122317196208507,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1309,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.42322878393402247,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1369,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.4161632577570773,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1293,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.43968697746070173,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1494,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.40193300260047393,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1235,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.4472572867814344,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1406,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.4080838561612705,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.136,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.39349700192036974,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1299,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.4244167932745341,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.1415,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.4250397190087772,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1405,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.4185301431892828,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1448,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.41837780099353444,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1422,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.40688623114654304,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1297,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.41152098100191425,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1325,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.43541616591951965,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1426,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.4239570313795913,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.1425,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.4179196132982898,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1348,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.40793896236230764,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1249,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.4067982423662688,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1219,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.43684860736228265,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1319,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.4291369133365572,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1475,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.3973133233125824,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1237,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.4364732674677368,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1587,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.40991302946633135,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1337,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.4146638735148003,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1311,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.40092817671934045,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1356,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.4353090605729608,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.1519,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.41133005090554353,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1071,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.4503627777363777,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.0762,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.39289736534749176,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.0691,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.34653932445666674,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.0618,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.33367564525962456,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.071,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.3111631450400017,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.0659,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.2719793832187608,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.0527,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.32460535440506744,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.0741,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.2941053593681007,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.0553,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.3963694072931197,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.0779,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.34352614429407285,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.0538,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.41474076174375074,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.0707,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.4277492204161283,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.0758,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.3489695902950081,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.0557,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.38468098451905947,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.0693,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.3340052894615738,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.0594,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.31203005467214634,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.0641,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.31253452672311643,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.0579,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.2956700245906162,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.0578,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.3094574822598369,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.0705,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.29328558836841634,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.0621,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.30485721996729964,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.0645,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.33338112629172284,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.0774,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.3170990833885196,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.0594,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.2893407158859673,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.0591,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.3123002699700789,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.0706,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.3263122774281356,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.0733,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.285597935864826,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.0583,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.3034994087411627,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.0619,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.2983343139244299,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.0586,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.3059278604856254,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.06,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.316248747282879,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.0636,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.31238165940355006,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.0566,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.3232884689171841,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.064,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.3051510649532108,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.0589,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.3367932615998432,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.0654,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.31445495588853944,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.0585,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.338557918974052,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.0665,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.32200877657285043,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.0642,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.3136652738730278,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.0657,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.3004037727919788,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.0622,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.31610736197278105,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.0703,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.2863970659579632,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.0631,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.28904088276145384,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.0602,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.3038701634909387,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.0648,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.2990377654173362,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.0649,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.3072796112430039,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.0702,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.2836915028039766,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.0569,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.2965990881021712,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.0648,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.302416505279951,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.0625,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.28286805849273106,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.0545,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.3175569782139587,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.0632,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.29192388643191347,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.0538,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.2950578761265139,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.058,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.3073529757920032,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.0612,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.31244742007920207,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.0725,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.3120032307724625,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.0619,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.32967249950399824,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.0652,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.31107968748574455,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.0665,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.2761236385855427,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.0595,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.3313365043783082,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.0748,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.2936517344634206,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.0556,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.3080441889999144,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.0593,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.28707294435180175,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.0529,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.3139781051760357,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.0605,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.33847380047044023,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.0754,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.3157934448109351,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.0616,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.27604450771769246,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.0488,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.30381137474227954,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.0608,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.3083108940971016,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.0721,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.2939969915768778,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.0587,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.2966260201125881,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.0652,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.3047388394646874,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.0621,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.28406299656767325,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.0547,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.3054448821369431,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.0592,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.31194745190925727,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.0653,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.27869803830710455,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.0538,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.28379619965085945,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.0602,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.3208477426644548,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.0697,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.3014293222919807,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.0656,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.3002851338231111,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.0632,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.3179908247796843,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.0653,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.3035797262185022,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.0648,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.29283007294405267,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.057,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.29901678896703116,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.0611,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.32601166189040603,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.0665,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.3015465385479113,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.0566,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.27323055527020856,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.0498,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.30730180548013464,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.0594,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.3146155692864183,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.0593,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.31410433267266946,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.0665,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.28752664109395054,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.0529,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.3207530078777094,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.0752,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.3078560983383755,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.063,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.3187718104110151,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.0689,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.28796456428663425,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.0592,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.3139692971506357,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.0719,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.31152967519225905,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.0675,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.29678992081069605,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.0553,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.29202368841583565,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.0566,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.308146398028009,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.0655,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.3268682733533068,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.0684,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.3074902364669794,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.0643,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.28477028645564983,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.0532,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.31305222243083586,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.0663,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.28566847502777815,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.0557,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.31917956522260305,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.063,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.2934369959853664,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.0593,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.31623648850021735,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.0619,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.2888202776235571,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.0553,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.31579110114447284,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.0649,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.33307754201536244,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.0675,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.2958187850520503,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.0534,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.29843792757480003,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.0539,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.3213869484375604,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.0648,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.2968949388784714,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.0576,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.3000010983609746,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.0577,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.28540193281177734,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.0533,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.32214500765186055,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.0649,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.29749776083692153,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.0639,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.28592026713698987,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.056,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.3128142552988372,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.0642,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.29031960119901945,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.0554,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.2867311487958154,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.0578,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.329563603506433,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.0785,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.27656517161431793,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.053,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.29921921397483325,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.0609,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.3116436857749491,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.0669,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.293367337690025,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.0573,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.3126583988454346,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.065,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.296818930286799,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.0625,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.30405529273560883,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.0639,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.32146677308978006,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.0679,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.28544422928401775,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.0522,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.4651103309359157,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.0627,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.3087922443199004,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.0632,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.3240407591957045,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.0688,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.2853978323079131,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.0553,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.30652737047049344,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.0607,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.28974446426274064,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.0593,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.3161376126367526,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.0665,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.2998304851503162,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.0593,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.3193834184274261,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.0654,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.29592421735799246,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.0632,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.3387743432994606,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.0643,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.29339273958599765,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.0587,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.30158184203495064,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.0595,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.31752353685161794,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.0581,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.28679170370095136,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.0543,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.3127053050066707,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.0668,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.2957412741521833,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.0623,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.3170222399102702,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.0682,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.29298038846458835,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.0605,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.29618327576337966,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.0524,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.325183518157176,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.0666,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.2907819598914672,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.0583,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.3231315356428015,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.0692,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.3189238356905663,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.0719,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.2825420997271948,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.0559,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.3054430064869817,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.0582,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.30452364771721696,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.0621,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.27923976452134824,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.0553,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.28341333288960446,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.0592,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.2804162411205903,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.0543,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.2995351688066895,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.061,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.3103864274550169,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.0613,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.30797941247312194,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.0626,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.28148358334476353,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.0533,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.3082937352235106,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.0589,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.314221889379343,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.0642,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.3102005392174038,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.0641,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.3011311189553844,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.0637,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.3218333585514734,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.07,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.27245527321445795,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.0503,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.30655775719811706,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.0652,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.31342514809450006,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.0607,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.29499260443633035,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.0541,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.30705444174420843,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.0638,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.3030472908868173,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.0613,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.2879047612319723,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.0566,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.2985827337178942,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.0569,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.31516534714480615,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.0765,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.292240809605871,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.0544,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.32499986410169135,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.0632,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.29212933820231046,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.0571,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.332626561188671,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.0754,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.27757486497553646,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.0506,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.28597791984256693,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.0611,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.2968513077505961,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.0592,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.2951607131276102,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.0556,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.2999680437153407,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.0581,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.3031614846584308,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.0645,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.3063492502587519,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.0644,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.29449640408200956,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.0596,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.3041720728808613,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.0604,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.3191614507047449,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.058,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.3056864042484999,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.0618,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.324885757169243,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.0702,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.31919957034919316,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.069,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.2779966824430725,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.0522,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.3115585703700135,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.059,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.2951435281049411,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.0564,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.29639644187911524,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.0557,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.2967059585533277,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.056,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.2962390977533624,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.0609,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.30772730731708425,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.0603,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.33167669569698205,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.0784,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.3031488614868127,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.0623,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.29004466471946794,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.0573,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.3418024583479107,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.0745,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.25944446858761916,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.0498,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.2991402692522587,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.0635,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.28382728142484004,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.0533,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.30013216511655444,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.06,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.31738962005013704,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.0666,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.27495811147987803,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.0534,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.2898001853575527,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.0587,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.3038253653042939,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.0662,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.31890592368762566,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.0712,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.30285297491583174,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.0643,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.30238437732364,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.0572,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.29919811505983684,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.0563,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.290310938842094,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.059,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.30406677284131756,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.0568,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.30274183530863613,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.0656,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.30846807084645106,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.0636,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.30277761626475513,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.0601,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.3026715769851036,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.0608,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.3022805304278271,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.0683,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.3000429195199446,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.0635,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.3000326393628918,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.0577,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.3054738242431294,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.0665,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.311565912227924,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.0639,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.30086679749542894,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.0654,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.30872146999776284,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.0631,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.2999008739434484,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.0637,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.2853655363231859,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.0556,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.3248617295559702,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.069,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.3025587502525917,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.0637,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.27976043161335323,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.0512,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.32672638946398497,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.0708,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.2925362124630575,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.0536,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.30465602129274244,
+      "learning_rate": 0.0,
+      "loss": 0.0649,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 9.629654902783345e+17,
+      "train_loss": 0.3744121859622438,
+      "train_runtime": 36395.1217,
+      "train_samples_per_second": 4.341,
+      "train_steps_per_second": 0.034
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.629654902783345e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}