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": 4.988354430379747,
+  "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": 10.597856579722624,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.9571,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 10.607043195525671,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 0.9676,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 10.706657444321783,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.9485,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 10.833775052111319,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 0.9651,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 10.50847582278368,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.9376,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 9.947414094766097,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.9128,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 7.499140471051322,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.8532,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 6.738809920806824,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.8285,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 4.013679344819486,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.7618,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 3.470424298308866,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.758,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 3.0857099909748364,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.7194,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.119722311619662,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.6722,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 3.8944096806977995,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.6618,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.5531547867367976,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.6671,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.2726158970866477,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.6509,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.7401841967387766,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.5685,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.524638726867235,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.5691,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.5366118787139773,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.5536,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.9339670880501139,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.5291,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.4837825629012822,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.5275,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.3571344813894985,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.5236,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.475000200534282,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.5074,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.1836485274849544,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.5122,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.1033110989064023,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.4691,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.0256121666546854,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.4702,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.1157658877688625,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.4668,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.029229025975574,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.457,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.9328508384383838,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.4522,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.0566112685471274,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.4506,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 1.1506745016862157,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.4406,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.8379116101607936,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.4335,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.8856371641708342,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.4153,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 1.040905725735143,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.4736,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8476214532428021,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.4359,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.8042108392207906,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.4344,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.9726086526039516,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.4452,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 1.005566970117761,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.4196,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7985274346790919,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.4249,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.8007233325482083,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.4061,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.7971653997776015,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.4128,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.7941021134368499,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.4217,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.7442312216878079,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.4092,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.856316447375084,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.4266,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.6919137993708384,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.4042,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.8236868534761969,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.4107,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.8447205346815115,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.4434,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7890575119240519,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.424,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.9623543479388135,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.3979,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.8872156887616599,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.4088,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.8435457130682839,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.43,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.8992344276562494,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.3905,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7866301068831886,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.4288,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.7647510596272384,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.4212,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.9147279600370907,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.3926,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.8771748604676273,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.4037,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.743479051191606,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.3969,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.8855391043808493,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.4322,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.7342537236301659,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.4145,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7107156394812874,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.3847,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.8852521149221663,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.3911,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.7654183064762566,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.4055,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.7725431735331209,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.3697,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.7671487854761218,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.4239,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.8256228716831271,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.4078,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.761009614427741,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.3943,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.7563723852911278,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.4012,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.7998896187505551,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.4115,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.8342735532859232,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.3952,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.8702733187409287,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.3938,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.8761324670334012,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.3943,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.8452018626348881,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.3975,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.7642538204924108,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.395,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.135099276215533,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.3962,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.7683707843542776,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.3938,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.832335430820908,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.3962,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.7394322810240933,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.3839,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.7628205532593477,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.3963,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.7644663918697698,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.4019,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7847155463331374,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.4069,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.1556891233190627,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.3855,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.7608304524317081,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.3949,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.9887072682706313,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.4019,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.0105482734463422,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.3848,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.8827395325731182,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.4065,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.7312531155965707,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.3946,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.8589751825142086,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.3856,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.8437412687937912,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.393,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.8112030467216289,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.3963,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.0304135200867068,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.4286,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.6956895341516328,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.3841,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.9029380600957265,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.3791,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.7220664679901379,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.3883,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.7552382094287886,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.3934,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.8813638044788806,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.368,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.724532415409025,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.3886,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.7676437480598177,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.3944,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.7881692638666778,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.3922,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.9061471447909996,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.3936,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.8670943671528207,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.3949,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.9259318501936648,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.3958,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.8030614765107318,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.3802,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.6931255470392292,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.3787,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.8187097216193147,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.3925,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.8424675965877743,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.3821,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.7851134432497692,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.399,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.793507961049737,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.3917,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.7760255003458156,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.3918,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.8460195768371164,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.3995,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.7930175532163505,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.3968,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.6423228850928051,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.3965,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.7470507350462195,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.41,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.6955721155921858,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.3869,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.7330146122636444,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.3746,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.7511932719099444,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.3864,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.6442299135175044,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.3814,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.7578775873303377,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.3911,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.8066707869605085,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.3784,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.8384785674445666,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.4104,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.7262960271084441,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.3812,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.7632627876610946,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.3837,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.8228927557007775,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.3911,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.7722259379165398,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.4045,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.8606614286879417,
+      "learning_rate": 4e-05,
+      "loss": 0.3919,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.7342445029824402,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.3775,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.8970397640875137,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.3884,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.7509111202796721,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.3828,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.7137160835463059,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.3696,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.7364313757816846,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.3963,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.6846898899199279,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.3955,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.6872660187607469,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.3769,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.6505277580819689,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.3705,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.7980097106299987,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.3648,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.7710654370952382,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.3959,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.7918480444559785,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.4025,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.7619931535570434,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.3708,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.7249054147103396,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.3829,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.9558351603932692,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.3751,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.3367898029782181,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.3855,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.6998561674039511,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.3776,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.880300674859364,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.3744,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.039548320871521,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.3898,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.6080642765519276,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.3761,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.6956735021030814,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.3942,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.8453558835761851,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.3832,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.8180635334319686,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.3726,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.6909435204852752,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.3795,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.9365901398192944,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.384,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.6800918461494655,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.3758,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.6575605764556519,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.3755,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.758692372309832,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.3825,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.7080588311451,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.3776,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.7108646275619653,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.3806,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.8512595222949317,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.3751,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.6145592641361397,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.379,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.6732048799327258,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.384,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.5388883869967647,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.3803,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.7147976276479225,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.3779,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7206288918495014,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.3957,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.5926504940240468,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.3807,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.6427563108100489,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.3796,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.7634818830219771,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.3874,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.6129578641854532,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.3553,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.5911606272917611,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.3825,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.5761162598095454,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.3881,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.6505149865136578,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.3871,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.6723583880025024,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.386,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.6762391017691768,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.3904,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.6416877079877742,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.3871,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.5563007800338655,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.3698,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.6517552513917253,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.3673,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.5433100503835695,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.3779,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.5746943660726466,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.3933,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.5933310550441021,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.3745,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.5917570486107221,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.3722,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.651044628944368,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.3753,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.6469322189252825,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.3736,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.5693144942737072,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.3697,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.6569756721417154,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.354,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.5379689964146745,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.3766,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.6945648423978326,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.3847,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.595508911716692,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.3876,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.5769293388835219,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.3834,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.6260163630010694,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.3645,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.5527743290045807,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.3951,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.5523034557172709,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.3766,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.6746154635510357,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.3745,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.5698622717869589,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.395,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.7217788523335992,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.402,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.6499537556172794,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.3828,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.752410517319556,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.369,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.652062718690699,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.3747,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.9000428399982677,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.3721,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.5889768583381754,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.3766,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.5957301767373852,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.3783,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.5474465748926499,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.3759,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.6719025758525509,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.393,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.5709066803361321,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.3708,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.5882047547403061,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.3633,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.6464343161222337,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.3837,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.5736136401126556,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.3856,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.6060970944472354,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.3815,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.6135086394397418,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.3805,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.5250134927780638,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.3583,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.5753863113017562,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.3674,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 1.0368442648943803,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.371,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6564150744032459,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.3587,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.8785654897905171,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.3687,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.6377220548345753,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.3743,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.5578881371282776,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.3737,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.8078827295383241,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.3752,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.5783782210769186,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.3649,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.7190780288552978,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.3803,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.6524997002857991,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.3777,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.6522861166183381,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.3719,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.681440087083467,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.3765,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.6647760614191143,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.3818,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.5197132995116226,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.3893,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.526643072539528,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.3644,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.53196986722736,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.3776,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.5149570857857609,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.3638,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.6625973881230639,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.3645,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.47479816434155786,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.3677,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.5546623306406324,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.364,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.5726333105715213,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.3784,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.503433838754135,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.3844,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.5627846225565724,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.3786,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.5487947258737063,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.3559,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.598506865334238,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.3677,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.47550010889765804,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.3597,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.6266674348904568,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.3714,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.5371638289334336,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.3475,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.6028984598888791,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.3632,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.5158460098392486,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.3712,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.680952867611393,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.3645,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.4557171846535887,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.3706,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.6901701745848734,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.3585,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6649722632159849,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.356,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.503538725395437,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.3684,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.61165661299005,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.3694,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.5326447722979604,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.3703,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.5573082260676383,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.3739,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.4764367445651524,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.3703,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.5030214106604096,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.3519,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.5115189974892786,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.3597,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.49505409743174905,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.3753,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.456984886353186,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.3553,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.5294601528625502,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.3198,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.5474835479617425,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.2994,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.4771361723582385,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.2973,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 0.6025644188884766,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.2956,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 0.5756966693285299,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.2857,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.44686498340925673,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.3097,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.5423916163187905,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.2943,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.5433660680700868,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.292,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.5396508788577122,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.2917,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 0.693347166553033,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.3199,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.5275139024834924,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.2978,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 0.7600876860550905,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.3219,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.5250697980824542,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.3098,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 0.6346503197807927,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.2795,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 0.49434109852575886,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.2833,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 0.6443236323520032,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.3027,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 0.4847957563530084,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.3029,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 0.7486312522164235,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.296,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 0.5406864538962192,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.3027,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 0.5940450308802133,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.3003,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 0.5918620582842639,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.3022,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 0.43903536859385656,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.2846,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 0.5815867992352372,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.3026,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 0.49363713794655806,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.2904,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 0.6219107623579282,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.2937,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.4842135272753768,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.3052,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 0.5550125901715669,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.297,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 0.45705951747299206,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.2908,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.5404317542965678,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.2915,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 0.43757718204182583,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.3057,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.4829452434873946,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.2795,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.5471537253692011,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.2861,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.42898701788020827,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.2877,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.5073519321156583,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.2992,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.49817547680218477,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.3023,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.6112212756387868,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.3024,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.5268075460456386,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.2976,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.5865108602743925,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.276,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.4837484983409821,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.3198,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5169611325439641,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.2975,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.586213420876021,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.2959,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.532174785995706,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.2865,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.4907407664753818,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.3152,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.5722182317036407,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.2897,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.449303946669289,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.2945,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.4892699817348566,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.3097,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.6471770579903214,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.299,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.48983785374587985,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.3075,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.5117268100241288,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.2914,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.5414849550166005,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.2831,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.48192963833292457,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.2988,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.5438503941913814,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.272,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.4347107624285219,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.3142,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.5059069288558368,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.3022,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.5155176879536929,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.3026,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.44376692856256295,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.3285,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.5312826688193825,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.2965,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.45937158020410673,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.3287,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.5082814086499544,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.3053,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.46829013902290206,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.2788,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.4527034702169429,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.2856,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.48687979335999587,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.2919,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.4586602810094919,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.3008,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.4844954484461281,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.3052,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.4814172035328755,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.3055,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.47886343934167186,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.2974,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.46874100381824496,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.3005,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.4386515166623982,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.2928,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.47281396486201294,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.2991,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.46190348293182476,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.2901,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.4521045815331591,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.3098,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.517853355710945,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.2909,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.48015156389415853,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.2982,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.4787086255805099,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.2847,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.49261845334460014,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.3087,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.42036596992582187,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.275,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.5280229585094462,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.3044,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.43889953077655264,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.306,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.5060222627760217,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.2951,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.4463206441927337,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.273,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.4929689842336444,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.3005,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.486152373067644,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.3153,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.4714280577949285,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.3008,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.44105860584492185,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.2937,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.5020215538144536,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.3071,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.4872512702698008,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.3134,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.5043628204413667,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.3107,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.4969944659278613,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.2981,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.4849749316168276,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.3102,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.5266479263848326,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.2983,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.4961302732756157,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.3066,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.4796359837273817,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.3029,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.4558872708719175,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.3027,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.4419678302275956,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.2828,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.4748852194320727,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.3017,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.42164032985195954,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.3049,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.427366437959805,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.3186,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.4691514512504033,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.2968,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.4353360657056754,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.3005,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.4521759666278227,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.3045,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.5518372856912764,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.3229,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.4757627467078189,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.2946,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.4387658543673278,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.2991,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.512670545999738,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.3014,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.45420980747100065,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.3105,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.5170223816370442,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.3007,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.5284683940138204,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.3051,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.4886278289586694,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.2969,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.5463936340886857,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.287,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.5123560255282911,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.3073,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.4281940468453028,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.287,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.44444195424735633,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.3032,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.44457647804788814,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.2964,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.4468991362533764,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.2992,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.4803428966475556,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.2984,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.5087205523856694,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.3097,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.4860062506052386,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.309,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.5082387055247037,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.2994,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5843651746295488,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.3194,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.4235036497738844,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.3102,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.5275027323035625,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.2943,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.40651157762161083,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.2731,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.4869057662108104,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.3037,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.5074510205800331,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.3131,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.4573380805359441,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.301,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.482950256036781,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.2951,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.43234637618149346,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.2962,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.489044419821727,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.3063,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.4532196503627039,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.3007,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.45651179128178065,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.2911,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.4269397426618019,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.291,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.4101963482309586,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.3064,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.45353426542712577,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.2948,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.44228199663509804,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.3025,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.4600731626054912,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.2898,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.4053755874171409,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.2755,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.4964880962818574,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.2975,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.40514877708646546,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.2979,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.47373142079651626,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.3077,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.4505815725274228,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.308,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.42155127508695855,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.3074,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.5109331385770844,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.3107,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.40824910525509944,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.3085,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.4678416009048589,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.3018,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.3969604854439712,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.2831,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.46299795864395515,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.2882,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.4237234631647839,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.3283,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.4358921686076,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.2795,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.4275525190819152,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.2964,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.47262055714563705,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.2863,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.45291425612281394,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.291,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 0.44688083119697897,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.281,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.4308197996011164,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.2881,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.45279113521632597,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.3195,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.4580672767901763,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.2979,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.43385303917334284,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.3005,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.4453842544165899,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.3026,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.44856694896444577,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.2938,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.4420839247960888,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.3012,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.45067997147780886,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.3051,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.4396303082790783,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.3111,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.46043760812229473,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.3046,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.5149389495930078,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.3065,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.4333416818786418,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.3174,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.424223100387683,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.2941,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.47155243801520175,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.31,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.4303129838855208,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.3209,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.4651601702867236,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.2981,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.4086257247443737,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.3007,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.45216041105742044,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.2929,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.4674620775465076,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.3038,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.38235284961307403,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.2858,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.43781155837567093,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.2895,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.4593083587928538,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.2915,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.39486199172004366,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.3069,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.5006146050890457,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.3231,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.4080132943526163,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.271,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.38612016130717614,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.2803,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.4859973389812529,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.3043,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.41418352093098143,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.2964,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.4021013042456477,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.2848,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.4874252424275696,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.2986,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.5066528769771336,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.3201,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.48561588171930276,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.2963,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.543192049545919,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.2858,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.47328667945012604,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.2978,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.4964437762555751,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.3136,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.5064132556485798,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.2886,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.43594300584762213,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.308,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.45710033127298877,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.3044,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.5120035307690651,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.2963,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.4478924625340222,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.3021,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.5424982898669775,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.3053,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.44658043637594114,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.31,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.48368253000343353,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.3128,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.4517929285940646,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.2811,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.4508049684639001,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.2693,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.46047314674057915,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.2892,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.5024883962098212,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.3059,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.42831528051468,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.306,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.46190420692011747,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.2959,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.3817642472424582,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.2866,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.44401059475878873,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.2912,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.374650123674571,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.2936,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.40481364712429796,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.3092,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.39922223581473953,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.2957,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.3958991268294826,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.2798,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.44843913371184846,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.2975,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.4334931856445464,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.3087,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.42583768756035656,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.2947,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.42338656853973783,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.3106,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.4402065939774769,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.2832,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.4083615099485341,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.2943,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.39319694269023736,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.2928,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.4792862997275552,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.3069,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.4750224314323678,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.2929,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.44203710238623023,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.3069,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.40730588536957313,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.2995,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.5036731618606649,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.3015,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.4487383684589304,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.2821,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.4220389714818781,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.287,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.5678021707477281,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.2928,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.4557563372132473,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.2903,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.4565702415540227,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.3049,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.4930551588640892,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.3193,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.3990548681332333,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.293,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.43883362331412856,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.3013,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.43835454655726475,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.3019,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.42333587527461264,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.3115,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.4104754709469071,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.3205,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.38885727568294093,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.2991,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.4612129224145806,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.2853,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.3948844201672649,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.2975,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.4165655028867134,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.3094,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.3968647942428107,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.3091,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.4085771765719873,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.2955,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.40885039437288656,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.3053,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.38752385347367835,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.2984,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.4334571365514193,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.3114,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.38434480898480383,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.3071,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.47365699646825965,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.3105,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.40142494813717516,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.3128,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.4633744797844157,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.3203,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.44114730127346236,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.3016,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.385963233434635,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.3044,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.4501627959162184,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3077,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.43241806228606433,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.2922,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 0.5265811109746119,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.2163,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.40957418072612595,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.2007,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 0.6506654580597403,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.2163,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 0.45753952104508816,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.1989,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.46747113666779533,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.2068,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.4524543224599547,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.2121,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.4353336358683354,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.1993,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.43725458152572627,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.2161,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.41383561136155506,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.2039,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.48394295428157286,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.2098,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.4659060092300586,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.2246,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.4588032693016991,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.2086,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.4474545479302261,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.204,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.5074816237146281,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.1961,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.4813184539807917,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.2247,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.49896929521037775,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.2124,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.45375319840860623,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.2192,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.45209802729826004,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.2057,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.4314757523237764,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.2085,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.4742391639522314,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.1991,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.44594727845121673,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.2106,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.49142284734745967,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.1999,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.4742823522113063,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.2003,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.47321193523374705,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.1912,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.42848914139384786,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.195,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.4289571099751011,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.2113,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.46216753702684266,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.2079,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.4304990560803078,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.2058,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.4423737868576725,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.214,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.4210909264742685,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.2129,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.39833206118247994,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.1923,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.41946283149149316,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.2066,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.39213313400001665,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.2053,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.4182927632413859,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.2083,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.39381586815633096,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.2002,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.4387523948414214,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.211,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.4090836423957057,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.1969,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.4240252154768084,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.2067,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.4840918546935191,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.1995,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.39160696647167026,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.1956,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.4677819395457515,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.1988,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.41261347324221587,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.2148,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.45659265290609485,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.1962,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.47746158851803594,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.2136,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.4839146892614473,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.2211,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.442653287014014,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.1986,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.409000501067148,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.1931,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.4688308168257384,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.2022,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.44229660080385425,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.2112,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.41014332343814414,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.2235,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.43426680344251967,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.2134,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.3870480356632664,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.1878,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.4297173719421829,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.2125,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.4226226376003894,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.2031,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.42202348281743746,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.213,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.3924862870803146,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.2087,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.41750285548086247,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.2162,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.45254972751199113,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.222,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.37790821052754053,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.2101,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.39535604942781644,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.1933,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.44689867715821435,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.2063,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.3909766964703794,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.2267,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.39944508270338824,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.1991,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.38976925245524585,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.1951,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.39355829604947246,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.2031,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.3890191096476927,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.2171,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.4120042953562287,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.1962,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.3829005659419243,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.209,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.41220428816384336,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.2054,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.3748760839242843,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.209,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.4364611201200056,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.2082,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.4208877770954868,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.2157,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.4270198112847584,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.1955,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.4326864163497178,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.2164,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.46052150558339794,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.2166,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.4041901769558889,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.2019,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.44453973895699955,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.2106,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.42294992587297464,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.2046,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.4406326685005854,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.1954,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.36734501630062416,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.2118,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.4517529272978518,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.2226,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.4254471494269942,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.2117,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.4322665299939197,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.1946,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.459859088576044,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.2116,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.4107739098879287,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.2065,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.4192469484374686,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.2006,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.4127854599935202,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.2243,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.41160737814753195,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.1993,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.3787132902967098,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.2114,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.4646538132249511,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.2016,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.4126588208066275,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.212,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.41913155034457794,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.2155,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.42528011892324247,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.1908,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.3621223990090238,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.1849,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.41933543691885683,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.2164,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.4011792015497497,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.2082,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.4121010116275741,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.215,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.41160319235127835,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.2041,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.3689237057061302,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.2134,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.44478708158567387,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.2152,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.3829465055435632,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.2196,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.3844726320845453,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.2213,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.43749074926830783,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.198,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.41815385842715475,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.2189,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.4159122784308772,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.2161,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.4093770050253238,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.2111,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.419047841843285,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.1998,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.416913736706924,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.2036,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.39263033480660064,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.2094,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.4294203370248884,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.2044,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.40416998451201075,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.2186,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.4315718570268212,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.2091,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.3778615875594799,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.2088,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.4119600736238594,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.1959,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.40975006239017925,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.2056,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.4141178057231586,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.2057,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.3783765914811497,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.2193,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.3720289108388669,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.1882,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.3882771469482746,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.1957,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.38763280860933125,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.1953,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.3901160268810308,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.2128,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.4119861439014156,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2133,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.40172314989764646,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.1996,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.3931307690386317,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.1961,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.38798053644931846,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.2095,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.4169126173133942,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.2153,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.3905789358620301,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.2061,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.4010515974935991,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.2106,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.3781325515440917,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.205,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.389473202336237,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.2138,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.3811910155163542,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.1944,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.41022235288733894,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.1967,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.42505459345035435,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.211,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.3739086460363547,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.2148,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.4187960552626965,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.2074,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.4033004988565843,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.216,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.3496523017533157,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.2025,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.41632257379558735,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.1886,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.3856683379153393,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.2069,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.3860644682753078,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.1968,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.39242657987636775,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.2186,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.39466281074785825,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.2013,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.399145796694383,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.2137,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.36985218024701216,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.2078,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.45565670910805073,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.2141,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.4037439314645099,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.2162,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.3883129396150831,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.202,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.44076060503250986,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.213,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.3846102049587694,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.2121,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.4001186475047884,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.2105,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.41221898396945206,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.2182,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.38256105570238663,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.2023,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.3829423593833692,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.2139,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.38601737313861545,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.1983,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.3718656326813687,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.21,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.39947470354510406,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.2106,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.3720717753605287,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.2147,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.4108015912210405,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.2092,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.40799750148451475,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.2053,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.37079417187178215,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.2039,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.3833553617237346,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.2155,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.38374990645506996,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.2053,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.394977263824125,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.2166,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.38974387067324806,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.2111,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.3777264363619921,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.1977,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.38303852619580303,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.2147,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.3969023602653192,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.1853,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.39068774763993125,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.2074,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.35651131332300484,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.1996,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.3698470995109885,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.1935,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 0.4168969623514211,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.2132,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.390209349219457,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.2045,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.3699734783933441,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.2156,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.35747485360872844,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.2008,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.37084250551807496,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.2151,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.3777257035563516,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.205,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.4032036998845211,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.2237,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.374527113298327,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.2072,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 0.3823339011776339,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.2029,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.3814174181842682,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.1995,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.42171735756056555,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.2226,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.4081794112122866,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.2115,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.40871244870657236,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.2204,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.4166827616143176,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.1998,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.43022600084173723,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.2046,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.42050330030986743,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.2121,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.449482793866975,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.2233,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.45119188598964677,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.2121,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.45186194351164477,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.2156,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.4354903628596846,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.2058,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.4072593579135999,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.2138,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.45202113688950574,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.2074,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.4515543409764128,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.2075,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.3984990199255621,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.2149,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.413067470413853,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.1972,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.429517598748894,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.2105,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.38809502611747476,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.1956,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.3778107438774699,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.2086,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.41666904000134436,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.2012,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.3943325558860255,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.2097,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.3897076086708847,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.2029,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.42061558937137694,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.191,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.37548387628795354,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.2138,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.42820137790645246,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.1967,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.4184813532057083,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.1858,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.362026259343848,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.2072,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.4403418687228894,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.2026,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.3610746999043713,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.2106,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.3765175272960478,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.2005,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.3985095663743469,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.1929,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.39919100185741385,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.1982,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.372838832730004,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.1915,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.37074456924488153,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.1963,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.3993173825813858,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.2209,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.357726062320682,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.1975,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.385130265925581,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.211,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.3639153507511063,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.2097,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.37305687225130746,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.2196,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.4066281621124897,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.191,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.3816916679729236,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.2335,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.3831402995919013,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.2049,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.4123876711191502,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.214,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.3789619284009373,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.2054,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.3951976401272903,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.208,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.38337779481867645,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.2097,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.38717635387089117,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.2204,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.3830072608010665,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.2113,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.3650985929554243,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.1943,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.3760339051245564,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.2102,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.36798294732103815,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.2151,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.3499742854566924,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.207,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.36424574038348084,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.2023,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.3817532229283645,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.2096,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.36759747966034007,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.2045,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.3941981967762186,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.2075,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.35243018308530355,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.1997,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.4191507804834559,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.2097,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.3724808915259928,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.2157,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.3756812341330994,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.2002,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.3813710500078145,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.2037,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.3991037706160741,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.2108,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.3613239324673112,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.2004,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.3756726258987165,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.1891,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.3963098390874544,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.1933,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.3785820583665009,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.1867,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.40516747956837695,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.2092,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.41390621411296397,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.1676,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.4078718448327497,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.1319,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.37453948154172406,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.1246,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.4788284572008087,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.1348,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.47183715128355613,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.1142,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.5155810219641521,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1437,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.40300684635949824,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.1154,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.43261123773018095,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1214,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.4609865476756988,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.1403,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.4104614893635468,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.1469,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.36148765666333804,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.137,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.42589321783386763,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1263,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.3908364027574495,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1151,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.37974475313030526,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1142,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.40739123845316905,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.146,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.37194214482450416,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1286,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.3660673482343212,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.1313,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.3488287697852377,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1323,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.3555131501829775,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.1092,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.3522344153251949,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1241,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.3505259590659311,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.1463,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.3647955682390095,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1366,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.34602891301402794,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1353,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.36046372020242246,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1123,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.3562849083597708,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1221,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.3431308659110261,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1358,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.3426604073966124,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1324,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.3774246491560965,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1332,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.3412512870387808,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.1337,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.34918383955276533,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1323,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.3538237693247829,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1079,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.3345278378524061,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1218,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.3559889590740548,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1217,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.36630708240740545,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.1158,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.34834197713850723,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1411,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.3612285208185491,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1309,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.3399670484896432,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1329,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.3588184249186849,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1297,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.3570458331130696,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1333,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.35409894739519954,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1334,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.36005317747434645,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1304,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.35392566740129505,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.139,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.36868459294941525,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1478,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.3679532011467689,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1306,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.37388529399355475,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1176,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.37120513657378423,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.129,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.353519662902085,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1085,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.3494633385268502,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.1444,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.3438533421104233,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1176,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.3736972641929468,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1297,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.32392113971325665,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1168,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.33954473745566277,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.124,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.3684666896589743,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1197,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.3569718648495322,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.1391,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.3299140077834563,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1132,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.33819434366567647,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.126,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.35739762251756185,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.143,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.3343976739280596,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1387,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.3536691518182843,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.1204,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.3429556516436086,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1279,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.3653330720786217,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.1243,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.3470247334502848,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.1275,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.3383095441969532,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1354,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.35466056502872595,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.141,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.344153493643548,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.1324,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.34064491759660476,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1364,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.34513145226087066,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1172,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.3471815805813307,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1201,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.34042451746544966,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.1185,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.33951165659281807,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1267,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.3331922073739917,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1192,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.3437845030489568,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1105,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.32951969792353053,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1188,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.35184359492611667,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.1252,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.36881887508439093,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.1321,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.3457492647892911,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1322,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.3403235385407729,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1321,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.3460707039039751,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.1111,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.3679146447314112,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1233,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.3386133146211636,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1306,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.35044379611500504,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1163,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.3546767879853252,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.127,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.34560033003579294,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1208,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.3402706890139215,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1263,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.343406833420776,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1268,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.34445492260049465,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1382,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.3285572252775729,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.132,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.35083480397119204,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1407,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.34011769067351777,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.1215,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.3467234690391032,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.1116,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.34155162161534725,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.1396,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.34956600192067105,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1381,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.33799842540439523,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.138,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.3463857926908405,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1099,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.3345615235399449,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1302,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.3739339652635965,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.1363,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.35426674846969824,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1202,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.34478947390574133,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.1399,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.3561288343924908,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1271,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.35512648420182363,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1166,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.36196971028272856,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1199,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.34612707725557595,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1364,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.3421378039981472,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1204,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.3198024910433534,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1241,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.34772354548514717,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1094,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.348737556216194,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.128,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.3270124424706831,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1227,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.3486382090742763,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1226,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.33732799399765123,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.109,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.34850988231629987,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1208,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.34578868192959855,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.1395,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.3218684846540958,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.128,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.3335724681610605,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1035,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.3463417060878717,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1447,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.3292504225831442,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1172,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.33670339257488285,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1167,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.3573243074733717,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.134,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.35791925567460403,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1216,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.36038221887187205,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1208,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.34109680331983067,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1165,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.3341777414054805,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.1204,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.3551518358483598,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1301,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.3506904296362331,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1288,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.33635028190575544,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1313,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.31776125055259097,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.122,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.3501837195385169,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1308,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.34547987419539394,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.1318,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.33528506496392263,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.129,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.3159583764520579,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1209,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.31891770788469104,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1223,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.32512901119219945,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1413,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.3335718312199352,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1295,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.33571726378150485,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.123,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.3267158737140349,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1115,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.33251548731888575,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1129,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.34562711245207817,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1176,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.3384481941582139,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.1182,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.3322348851605928,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1316,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.32778218011268473,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1296,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.32155014577220337,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.1412,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.33167142874823374,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1196,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.345996714535005,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.124,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.3321182045796083,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1285,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.340658085311028,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.1445,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.32358706248560115,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.1239,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.32532006330378477,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1323,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.3483253778201618,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1351,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.3359050422577944,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1233,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.3325930739443392,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.129,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.33453628919150574,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1272,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.3296998606769133,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1174,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.33738893670259346,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1327,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.33557106336833337,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1116,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.32730761308740297,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1271,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.3281391283586037,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1274,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.33024850126702554,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1181,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.33145242930945845,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1356,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.3286776326061526,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1265,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.3203830865233062,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.123,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.3237374680317796,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.1261,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.3231152355761042,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1277,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.34736475303470943,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1341,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.32438430293802883,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1375,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.31352600145851156,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.1314,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.32646469770722136,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1116,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.33584427414090906,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1263,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.3296920133076922,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1227,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.3270178085013252,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1259,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.31932560592890175,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1148,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.3409024576961353,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1322,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.3149730029495649,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.132,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.3201698590997085,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1277,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.32020220143852346,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1154,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.3316251424921194,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1186,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.33245124978223883,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1241,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.318570183976669,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1202,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.32653965090584375,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1172,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.33651093834132906,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1053,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.3293269182509134,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1153,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.3321711292815959,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.137,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.3279843436080548,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.1297,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.319400666254999,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1319,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.32547178623958345,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1225,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.3325906795318583,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1206,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.32019109635549126,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1181,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.3076931770183904,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1254,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.3244418426175658,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1161,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.31155745109754973,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1182,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.3335515793475797,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1414,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.3279426431791744,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1324,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.3422518146311021,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.1228,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.31262038811834275,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1092,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.32544218281607323,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1426,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.3297023655390643,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1315,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.3193210201824807,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1252,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.32076211342960337,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1151,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.31733621485545743,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1115,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.33557833890040983,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.109,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.3265141263132453,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1229,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.33695264884440257,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1266,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.32201603860747335,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.1301,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.30243254396673347,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1326,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.3228422824947193,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1311,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.31650768302917687,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1237,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.34366562184560834,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.1304,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.33140790639463963,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1356,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.3099267400844887,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1138,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.31662882764525535,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1302,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.3109049593441237,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1175,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.321811115169621,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1148,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.33423209616405875,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.122,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.3235397011517906,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1184,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.31628234448433773,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.12,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.33053111648991784,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1111,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.32242569281261213,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1165,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.3124873924114078,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1291,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.3233381682255521,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1136,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.32717964160328833,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1132,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.3181663712420808,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1211,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.31969481234425073,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.1262,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.32049726691153746,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.1311,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.32147117417586707,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1202,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.31575171131631974,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1152,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.3111711666507587,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1193,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.31579995881850353,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1223,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.31219005432509334,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1348,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.316326250806724,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1308,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.3236749685923383,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1228,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.3383079141246945,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.119,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.31424034413959345,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.1237,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.3202987164351996,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1211,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.3119038250235825,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1147,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.3263966130728821,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1207,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.3265045992453955,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1127,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.3220485094693995,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1438,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.3242596991408637,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1178,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.3250781494559231,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.1227,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.30728224798151427,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1272,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.32704004028428646,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1143,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.31807415859768323,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1232,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.33648556619857917,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1171,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.33807513342089074,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1268,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.32803269206775,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1398,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.31587351731928853,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1019,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.30546888114372545,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1282,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.3293802513476139,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1374,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.31753410412700545,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1255,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.37919928053043167,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.0771,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.3153040419505681,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.0667,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.28934087920211715,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.0776,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.2546419981353757,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.0836,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.27111712161177537,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.08,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.2891444022123417,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.075,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.3615080699151879,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.0849,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.3634718103885271,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.0748,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.35270187750304266,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.0922,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.3136301496912693,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.0741,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.2908952561270445,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.0805,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.2946866214321842,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.0893,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.31068934885645244,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.0813,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.3029509689870957,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.0965,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.3269601374341932,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.078,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.29277874942160115,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.0754,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.27185938071018567,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.0774,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.29134303610849666,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.0795,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.25676530335120473,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.0762,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.2974454078961463,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.0951,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.29218647503921413,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.0765,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.2854702760354743,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.0743,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.2738924976124304,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.0752,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.26080483989137515,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.079,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.27511823097111654,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.076,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.2608781815288883,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.0814,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.25214521757260355,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.0735,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.2621047848357154,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.0724,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.2629980658009185,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.0896,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.2722389779990222,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.0729,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.26979230030706686,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.0797,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.2659467847700565,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.0867,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.2625574099163564,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.0565,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.27302507415763894,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.0725,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.2658449851467123,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.0747,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.2747421676947847,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.0709,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.27872716996147395,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.0844,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.27413497793283137,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.0722,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.26351313864162845,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.0754,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.251101142248738,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.074,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.2686851283571928,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.0935,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.26332592442709474,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.0767,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.27945787751835593,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.0808,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.25232662578972564,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.093,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.2523944341017427,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.0797,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.26503986584478334,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.0725,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.2513135001538848,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.0825,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.2552876992805431,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.0701,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.2596382009871985,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.0753,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.27781628859128227,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.076,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.2716958088618469,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.0754,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.26290788999985587,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.0802,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.263229956609127,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.0674,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.2726634184409356,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.0773,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.2580930413053273,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.0765,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.25492581870358844,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.0681,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.2573635708409566,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.078,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.2625416659529986,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.0863,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.24101408267915217,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.0776,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.2556396446887368,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.0816,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.2661678023058352,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.0831,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.26274759337868303,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.0777,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.2539375445682434,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.0731,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.26219225101700067,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.0868,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.26831650525809925,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.0851,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.25400193592601494,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.0682,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.26104004396906844,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.0854,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.2524704197077927,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.0832,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.2655391518270285,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.0881,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.25990034723866073,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.0719,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.2717522029864333,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.0773,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.25158619337293153,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.0759,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.2647971117845565,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.0701,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.26167876062489087,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.0783,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.2608371501438563,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.0553,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.25744203769621476,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.0724,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.26044648063934056,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.0735,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.2548924672759011,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.066,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.2616289499485263,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.0761,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.25013735346096894,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.0777,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.2624642607504981,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.0687,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.2554697116666296,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.0761,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.25755028578274347,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.0838,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.2508551325136935,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.0926,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.25517738499228915,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1002,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.2592686965229478,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.0783,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.2544564071756488,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.0779,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.25858082449120573,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.0652,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.2513279834977404,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.0825,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.2705195645929642,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.0762,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.26911987614988775,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.0796,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.24473270241168324,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.0779,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.2477650679330589,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.0784,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.2649587594234352,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.0979,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.2652067174870674,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.0691,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.25873059043309643,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.0768,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.26172768339179786,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.076,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.24874556810594556,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.0675,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.269897631766143,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.0816,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.26036423782340834,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.087,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.2705472751865011,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.0911,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.2579990191860329,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.0689,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.2563501273008215,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.0714,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.25266530671900594,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.0791,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.2684824193982774,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.0724,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.2756362117964069,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.0637,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.26270870388464657,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.0635,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.2508727153986751,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.074,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.2603742573929357,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.0788,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.25272772359316953,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.0826,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.24865777017209884,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.0663,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.2580007950437733,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.0846,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.2612530461607611,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.0714,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.2539593490654387,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.0681,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.25713834690469917,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.0827,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.26861345189356406,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.0747,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.2612975293296721,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.0784,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.25600793352535467,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.079,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.25283112677230496,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.095,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.2702735589916005,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.0839,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.26178870366185913,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.0705,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.25783537133960766,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.0823,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.25280067107128323,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.0784,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.24772433419063422,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.0714,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.2675614874461402,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.0876,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.248467907327836,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.0847,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.24728266488504758,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.0796,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.25333574776964707,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.0807,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.25718205031923397,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.0894,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.2524479803668579,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.0856,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.27537586499668965,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.0707,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.2607596191911143,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.0688,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.24967770990299823,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.0882,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.2526295995921072,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.0727,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.26353165981942206,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.0845,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.25836846523780294,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.0781,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.267672087650614,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.0857,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.2718827432999498,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.0719,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.2597529146084574,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.0871,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.2549640948701708,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.0802,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.2658656218426949,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.064,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.2525825806525175,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.0835,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.2622457725345039,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.0835,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.2611488052615112,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.0768,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.2584250979708258,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.086,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.2740830600065115,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.0781,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.25562801899084103,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.0692,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.25516404530185555,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.0681,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.2485247908135712,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.0802,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.25098966124240274,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.0752,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.2418020728342686,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.0695,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.26126042178200043,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.062,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.2540916066305588,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.0748,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.24811957572345442,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.063,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.26503012878665516,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.0844,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.24715220538833665,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.073,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.24020423737902646,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.0729,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.25481714096241076,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.0675,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.2527412119832004,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.066,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.2623683494876068,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.0764,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.2804805539308191,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.0852,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.24916392678340843,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.0763,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.23836765240674818,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.066,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.2564609626120107,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.077,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.27695657601146767,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.0866,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.25466180044305287,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.0836,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.24654692578593293,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.0901,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.27535014541365027,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.0667,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.24716716777628947,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.0758,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.26152872785925513,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.0759,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.26978555768845525,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.0867,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.25654618105304516,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.0801,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.263845724438766,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.0715,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.25140750821045427,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.0713,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.25592122915993193,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.0669,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.2549247613415155,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.0792,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.26836934445143845,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.067,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.27135632141813326,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.0828,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.2529250188522654,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.0868,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.2525558403443637,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.0719,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.26709414015211114,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.0826,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.24395547528919392,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.0888,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.2497946008180024,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.0632,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.24813051425564867,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.0828,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.2564011714192998,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.0734,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.25550321477706134,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.0785,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.25330848781792514,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.0828,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.26415017571681193,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.0744,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.2571251873830965,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.068,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.26309024427289696,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.0763,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.2483204909367807,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.0687,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.2531355178116246,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.0901,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.25902809282289835,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.0701,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.2646293953092241,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.08,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.26208131472072127,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.0886,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.2500921326827395,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.0692,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.25665238310235844,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.0641,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.242277881184426,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.0705,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.2539199959463628,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.0703,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.2525211448279266,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.0743,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.256228301565555,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.0741,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.25241772063490986,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.0793,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.2543400221957755,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.0844,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.25710025334250136,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.0789,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.25450813602112526,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.0981,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.23990177655631478,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.0676,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.266177725216914,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.0688,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.2533481092171896,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.073,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.27014566811555674,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.08,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.25282280505965504,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.0772,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.2487377276353882,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.0681,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.24376373665779016,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.0704,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.25782364966411553,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.0856,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.24587899119347503,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.0623,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.250278699001532,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.0643,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.23949819347027465,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.0691,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.25521485154519447,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.0661,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.26518498455102857,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.0906,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.2664930934749687,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.0765,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.2465081991450239,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.0708,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.2650273638895782,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.0679,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.260450678373038,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.0773,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.24044519025420175,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.08,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.26063348646994583,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.0671,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.26322642043369227,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.074,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.25953952923965173,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.0739,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.2520070431252204,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.0872,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.2549765590315624,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.0799,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.25901039964079237,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.0678,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.24906571440025219,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.0909,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.2652648938871328,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.075,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.25769567062377335,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.0649,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.2621208152995241,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.0686,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.2575548428623965,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.0727,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.3007556311684188,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.0809,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.2513772292801896,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.0795,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.2751285422879927,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.0722,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.27166544970228956,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.083,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.257224101619337,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.0865,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.2564462095729044,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.0756,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.247912357387905,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.0742,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.22973030074657746,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.0798,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.2485185841873948,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.0677,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.2633033100639795,
+      "learning_rate": 0.0,
+      "loss": 0.081,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 1.0144059786589635e+18,
+      "train_loss": 0.22610286643834618,
+      "train_runtime": 38049.797,
+      "train_samples_per_second": 4.152,
+      "train_steps_per_second": 0.032
+    }
+  ],
+  "logging_steps": 1,
+  "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": 1.0144059786589635e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}