diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.982278481012658,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 5.450702942839219,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.112,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 5.369417163813169,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.1064,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 5.405520659758402,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.1007,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 4.90933358970897,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.0489,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 5.326313809453253,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.1336,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 4.876963145502961,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.0711,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.127562141348089,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.0497,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 3.699632931373633,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.085,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.11792961965526,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.0013,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 1.9184390267913694,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.9872,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 1.8355297671974844,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.9796,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 2.716683226021473,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9432,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 2.850458624314029,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.9612,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 2.6033071340982574,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.9137,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 2.3718914155543818,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9019,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 1.8329046351641762,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.8951,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 1.9204485642038622,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.8719,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 1.7658915710311622,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.8853,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.5915106629934486,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.8565,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.3056624514999797,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.8582,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.001516514885975,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.8531,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.1152907802636833,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.8593,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.111856883220982,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.8316,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 0.9649186889538367,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.851,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.9071543108422183,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.7979,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.8570884889081569,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.7976,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.9639458172904218,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.8321,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.8589334969312636,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.7978,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.807159810992645,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8173,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.7646691460431899,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.7817,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.7828711051482595,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.8188,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.8208679901259747,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.8148,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.734853147583619,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.7739,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.6745142232564065,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.7835,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.7042435894639656,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.7911,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.682661838153839,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.7929,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.6741616241193031,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.7848,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7250073494046341,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.7947,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6975975499394308,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.7333,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.6344450688747495,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.7458,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.6894685164919552,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7587,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.6007401322566721,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.7362,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.6347887396958207,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.74,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.5940643320102239,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.743,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.6161278829564315,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.6805,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.7140871852655967,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.7761,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.655981019171505,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.7779,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.6182862740882721,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.7256,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.6233094352047821,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.7243,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.7155973608362316,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.7518,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.6763718633980496,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.7554,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.6545248585148763,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.7388,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.6373112444527651,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.7266,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.6676047495303983,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.7292,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.5869262372845208,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.7626,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.6366629225776075,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.7216,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.6415095922785259,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.7354,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.6892525870403859,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.7264,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.6579883460768402,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7659,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.608901550243272,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.7082,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.6224576740407995,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.7333,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.7426624395974998,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.7063,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.7962223345660973,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.7281,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.6786129250973855,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.7393,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.7009393140461145,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.7019,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.6683001378729063,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.7431,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.6494993065257653,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.728,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.6292632157917815,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.6763,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.6287739336301869,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.6904,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.6309445186662429,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.715,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.6865005393917105,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.7196,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.6939463323779541,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.7427,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.7502140420326471,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.7524,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.7257837929522553,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.7099,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.588380118464532,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.6607,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.8284491245332587,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.7058,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.7317828171947689,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.7132,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.8540286929112156,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.723,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8478577585502987,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.7036,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.8762226488322865,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7073,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.8169041861786382,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.7283,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.7510436026740966,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.701,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.8108433931839271,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.7365,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.738255749526459,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.7006,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.8249630938814416,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.712,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.6777725981771855,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.6789,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.7271216782003657,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.7137,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.7276578833080114,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.7153,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.7618825259490798,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.6955,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.6314413900508642,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.6569,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.7714193990954729,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.7508,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.6653528908423856,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.673,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.6991659135869854,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.6988,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.7391232443717163,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.6962,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.8032151743145193,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6543,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.696625446546452,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.6865,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.8851168426027228,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7051,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.7062530042060008,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.6641,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.7960608651322016,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.69,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.7432476105463828,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.7115,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.7824868080372209,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.6975,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.738837065671878,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.6897,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.7470527353532169,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7003,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.9192361818251373,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.7052,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.9445960287209855,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.7026,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.7774273556546372,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.7151,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.8132048932604548,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.7115,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.7756239657920906,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.6677,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.7801748661646215,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.6828,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.6610452428490621,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.6722,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.8313477801310054,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.7153,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.8328063404214664,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.7323,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.7481791235285739,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.7314,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.7142039104568193,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6968,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.7180159790834836,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.6531,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.7274414030796118,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.6873,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.7516591731448778,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.6608,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.6785126949871667,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6923,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.7730305564115398,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7063,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.7946049617530072,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.705,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.8185728442531599,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7214,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.7995768553089546,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.6844,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.2077961644233817,
+      "learning_rate": 4e-05,
+      "loss": 0.6882,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.6760754290259714,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.7243,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.1182093367254646,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7107,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.8170228891689681,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7111,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.8355997220039678,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.7106,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.7424399798384652,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.6908,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.7261458665408289,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7158,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.6442793104667625,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.6978,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.6646604427594942,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.6784,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.6237923021948718,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.7283,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.6878586377854015,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.6863,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.638862146435286,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.7039,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.6935821280748589,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7072,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.7558268086461658,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.7132,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.6177609497998109,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.7086,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.7455930787181975,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.7281,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.5616031718832146,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.6353,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.6469137917800917,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.6986,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.8083323603447969,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.6962,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.920971764814151,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.7647,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.69788122574487,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.6769,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.6914140141242198,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.6756,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.7855610900015265,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7042,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.7438447329853578,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.7565,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.6791241155449038,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.7139,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.7560348355801471,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.7166,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.7933204492973033,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.692,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.6126923239378588,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.6592,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.6615944593716592,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.7087,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.7560297446909378,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.6786,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.7256272407785167,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.7059,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.7486088682626854,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6576,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.6701347926402963,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.6869,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.6299430805311508,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.6798,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.6761283896240854,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.7048,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7595534148788751,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.717,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.7133436998882573,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.6622,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.6629291488427469,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.6716,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.7628992694647982,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.6811,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.9604952194074989,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.6539,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.6202882652879521,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.697,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.9246044575969158,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6618,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.6134223357195765,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.6755,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.8828361856722766,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.6583,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.6070989057233199,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.6714,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.9417780656184999,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.6795,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.6626067925269281,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.683,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.745384240173719,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.6321,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.6789927558714048,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.6685,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.7377615407038279,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.6965,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.8514647488034012,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7202,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.5834157414042498,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.6618,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.7360654126509235,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.6735,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.6469021318147177,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.7136,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.7675181777351622,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.6923,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.6048176682721526,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.6656,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.838319993725958,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.6854,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.6744827708945915,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.6958,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.7408851053294587,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.6876,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.6670883244034604,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.6747,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.724773362125924,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.7123,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.6991330420002562,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6388,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.6805469546266123,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6804,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.6386639714990481,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.693,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.6205097322280176,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.6701,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.7237277097445686,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.7024,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.6553454240080242,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.679,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.5800568108047673,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.6473,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.6357745358741353,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.6545,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.6278589911495068,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.6785,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.6675247901928532,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.6655,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6508264592628574,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.6794,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.617547090383432,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.6682,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.8085832348473719,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6856,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.6726241883379507,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.683,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6348291965764816,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.666,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.7435099019042328,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.659,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.753852362281321,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.7278,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.7861398394115516,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.6925,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.6307619470290331,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.722,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.7578719877984802,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.6521,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.7848054158919294,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.7071,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.7491748089662942,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7012,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.7345683710307003,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6705,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.613646875303154,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.6862,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.8388629400691655,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.6321,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.5752584399974031,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6862,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.7389661194387793,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.6143,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.590620183309574,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.6562,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.6741080608817207,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.6587,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.6583047098617182,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.6526,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.6126874352237581,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.6919,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.6962626734074606,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.6584,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.5984304377803882,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.666,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.666115119029744,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.6504,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.6615742696659571,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.6862,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.7685336136914734,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.6834,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.7823188190256627,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.6629,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.6361359429915945,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.6536,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.8078862957719508,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.6761,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.5768025107682712,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.6776,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.702531608311987,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.6972,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6213198390408323,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.63,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.6248628722286187,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.6972,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.7955057985198207,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.6906,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.6062754177577451,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.6923,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.6907026082779005,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.6501,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.5826222087144555,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.6296,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.7308588783638498,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.659,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.5809033400863094,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.6606,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.733467032331028,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.7165,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.6094706329449876,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.6482,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.7348613053400639,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.6299,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.633972640924958,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.6823,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7459244807986091,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.6625,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.5671672015353476,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.6476,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.7743015269171095,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.6482,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.6487212204201589,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.6523,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.7008086821447416,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.6871,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.6846669336418171,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.6749,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.5878422139600205,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.6599,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6451556542977616,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.6666,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.6448155962563349,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.6578,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.6876682532738556,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6947,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.6995631248258093,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.7203,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.988912985927034,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.554,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.6829514983026761,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.5068,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.8547256873296442,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.4922,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.6584856874496007,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.4668,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.7378113951519617,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.4768,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.7008061897367528,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.4781,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.6967759201206042,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.4789,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.6688534644970603,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5311,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.6837398861520889,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.498,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.6186266909853625,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.4992,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.5785667005095929,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.4676,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.6179654379372564,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5063,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.6217014748257197,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5068,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 1.6554362273421828,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5134,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.6059081397637561,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.4803,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.6574364585233664,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.478,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.6510000461028382,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5156,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.7801956884064966,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.4831,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.6050657454065949,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.4919,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.6125339575812948,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.4764,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.6571186732103441,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5152,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 1.105699296324299,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5204,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.599497633826312,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.4939,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.7001658362968455,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.4909,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.8242967241587905,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5054,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.6075995965657577,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.495,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.7689188361757813,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.4864,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.6723454753580747,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.4858,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.6232571456600922,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5088,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.6865575712308168,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.47,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.6262558742854332,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.4976,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.6041766055652201,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.4874,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.678340519247428,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5503,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.6011391835820683,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.4891,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.5469862976998046,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.4673,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.5948402289260246,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.4781,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.5527120757214277,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.4666,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.6397025283845145,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.4977,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.5881221403315967,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.4752,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.5515399580348613,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5127,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.7573267826365883,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.514,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.6101696270164838,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5137,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.6558136499909372,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.468,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.6561794831874128,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.4879,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.6521245672013325,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.4983,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.6532860955550995,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5036,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.6523524007294017,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5018,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.6238092643727698,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.4962,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.6582317352471276,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5113,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.5878194633945559,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.4759,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.6707335167678224,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.4825,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.5245096396116243,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.4943,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.6643099698029667,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.4916,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.5664957193966443,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5131,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.6116469669186629,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.4971,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.6369321812564063,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5012,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.5804329554349709,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5208,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.5728309437042937,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5072,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.6527637160522138,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5097,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.5427909906970412,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.4715,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.5208681624952771,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.4601,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.5766726065996345,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.4948,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.5112766376185188,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.4719,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.5190600837702004,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.4875,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.5858709857405863,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5035,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.5503825758247286,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5032,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.5377746640037303,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.4838,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.5966497935721311,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.537,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5066879249906735,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.4761,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.5143595280532314,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5018,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.5749552103574224,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5007,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.5426647951396482,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.4939,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.510718898456341,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.4674,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.6374834420975903,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5041,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.5582692593874311,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.4842,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.6519620106147035,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.4658,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.5830831187358175,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.4616,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.6507324877137497,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.4759,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.5416668010859917,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.4689,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.5897961656292687,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.4761,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.5210302937952241,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.4983,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.6100845966874855,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.4917,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.5516596962053174,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5077,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.6034313520672853,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.4929,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.5627778429316652,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.5093,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.6115907736929062,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.4916,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.5405707410033432,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5141,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.5930998274034384,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.4731,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.5069968096232118,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5019,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.5181652843679097,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.4938,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.5903811013457665,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.502,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.5237273116305593,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5005,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.5974371435975578,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.4894,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.5289257841489882,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5038,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.6028324874172387,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.4954,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.5969996901240513,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.4983,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.533706421606113,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.52,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.6205441737911033,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.496,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.6587606543568972,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5279,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.6097958406638765,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5027,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.6515488828931433,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.4818,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.5866713684744517,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.51,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.6847828462100085,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5121,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.60105790552141,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.518,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.6787048255940437,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.515,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.5803937716795676,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5119,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.697721129762596,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.4957,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.5509741276782784,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.4724,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.5901909442305094,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.4653,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.5242853076806266,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5102,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.5894010047834797,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5105,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.5320637586798072,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.4942,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.6014970513805916,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.4772,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.6092850379008897,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.4962,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.7030872673088895,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5076,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.6122820809810477,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5148,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.5532714641164153,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.4718,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.6951250637345976,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.4751,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.5769948522624311,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5194,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.6487703457062757,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.4608,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.6211618727796604,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5048,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.6383124607840136,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.476,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.6074082200896511,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.4847,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.6506001507901732,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.484,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.5316122050669067,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.4988,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.699221261364724,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.4985,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.6006569095478685,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.534,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.625989083943264,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.4918,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.5701174945857882,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.4842,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.5547791097247635,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.5102,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.6120493154326229,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5073,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.569831131016809,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5008,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.5670448322507695,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.531,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.5641092404150295,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.4749,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.6153002519377766,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5274,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.5455713817254649,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.507,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.6188974243142922,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5156,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.5340602158069462,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.4999,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.5794474856786515,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5161,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.5640474611293268,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.4929,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.6060997413822471,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.4808,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.5440512916431063,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.4953,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.5892128004625793,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5353,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.5305696803236576,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.4697,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.5489049221621667,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.4965,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.5344467553099754,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.5086,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.5829894166309476,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5108,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5289960437440904,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5069,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.5269761132016408,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5433,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.5287724607674017,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.512,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.49348650022582913,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.4628,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.528408586326906,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5182,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.5288554757859395,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5013,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.579071133450241,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5103,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.5416571080456081,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5351,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.5590232241456932,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5018,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.5483803278735366,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5278,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.555206833043853,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.4825,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.5548055634930987,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5274,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.53116601653616,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.4866,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.6437954020521071,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5281,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.5583154815425904,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.4759,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.5796387600727607,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.4812,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.6033781155191976,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.4855,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.8288124522954978,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5201,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.5293122149310403,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.4912,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.5755986319256328,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.5842259735482,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.538,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.5150839653141213,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.506,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.5627537990561252,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.5314,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5628669216652122,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.505,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.5716749923497189,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5117,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.5505635936075719,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.4973,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.5678365662723992,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5018,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.6345534592848004,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5187,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.526644911454519,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.4937,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.6057981495607384,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5026,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.5461092552992604,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.485,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.6201628799946792,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5206,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.5101261590974714,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.5162,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.5482556083183315,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5176,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.47753590861496875,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.4819,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.4966254377716573,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.4596,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.5853453504941586,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5393,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.4640854113470312,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.4793,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.5584221939049504,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.4914,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.5018621927400404,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.4655,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.4505580803394948,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.4481,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.5472664499932293,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5393,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.4746854232880266,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.4793,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.48426884277829096,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5085,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.5042169825622934,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5001,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.5326850689030131,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5714,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.5720223125403432,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.4788,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.5291892955532951,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5149,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.5590338835150369,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.4678,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.5585340651610905,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.497,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.6447378316671825,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.4762,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.5182127916860229,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.4955,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.6400808021582163,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5044,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.5109411802955469,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.4671,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.6246893205631904,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.4808,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.5598998815657176,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.4686,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.5403062004137897,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.4816,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.6168126556409969,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.494,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.5615467678918586,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.4949,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.5576555175383344,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.4904,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.590619232816714,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5232,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.5366967402176797,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5157,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.5298851005131622,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.4905,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.6015259021275448,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5059,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.5000380207572371,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.4992,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.6201348418526236,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5239,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.5211905688235091,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.4658,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.5612186372529234,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.4843,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.5698226328355931,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5162,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.519988741647798,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.4805,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.5420128477457138,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5322,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.5202978370988293,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5054,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.5474096061186704,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.4912,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.4908778110646172,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.483,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.5330276648812251,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.4585,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.5202974366251839,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.4755,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.55081624696567,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.4933,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.5358569196006315,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.521,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.5844895695397885,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5242,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5461007632083009,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5486,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.6018226739085878,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5174,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.4914337822098023,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.4891,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.6212779781477425,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.528,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.5501903669489395,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5307,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.521775746742169,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.4701,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.49992072425231104,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.472,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.5653150419231634,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.4822,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.520047535171642,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5212,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.4952580222783512,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.4674,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.5016689872989418,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5097,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.5663917768566252,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.4976,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.5100064948416977,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5034,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.48642373846912207,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.4726,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.582254617058578,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.4988,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.4945453329513595,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.4884,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.5682214062455839,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.4711,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.538501178546766,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5196,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.5552808893757925,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4726,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.5000814759231574,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.476,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.8218873923968989,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.5188,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.7804791166139983,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3037,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.6757620137547422,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.299,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.968257302300952,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3089,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.6086297168689698,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3008,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.5384495387411458,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.2951,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.5893107798053144,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.2784,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.5822069740253141,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.2887,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.48032375581705083,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.2498,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.6051430928840648,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.2955,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.5283844994814993,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.2857,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.5598121174369511,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3005,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.48791450063937447,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.2811,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.48348039364363554,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.2619,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.5122651643158086,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.2721,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.7354077559811799,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3545,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.48577956161585345,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.286,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.5763778825019756,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.2836,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.5057773544536843,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.2911,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.5136766349548332,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.2893,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.4732465694888309,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.2605,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.4935291747001262,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.2803,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.4948517128590334,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.2986,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.4893917711721737,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.2858,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.4698681580021272,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.2753,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.5179194967889442,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3388,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.5056168452566384,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.2994,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.48499168527642117,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.2758,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.5151950842720042,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.2609,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.49588154859403566,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.2907,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.4900678122961883,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.2963,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.5259458398659657,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.2893,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.4307387227534818,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.2699,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.6255792171613523,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.2875,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.48538023457839363,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.2816,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.4533735710672914,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.2576,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.5429836308707117,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3009,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.4639586404015918,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.2725,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.544117076675063,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.2964,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.49227623842895274,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.2772,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.4931108879710274,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.2978,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.503408219965575,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.2944,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.48798917447208123,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.2791,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.5139300368853376,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.2924,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.5314768816394192,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3001,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.5104984844308839,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.2619,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.5028369459105179,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.2781,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.5478130895025287,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.2985,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.4930867904520722,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.295,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.49088875344629923,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.2768,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.4931276992373915,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.2987,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.4579779057493062,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.2692,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.5164928356358643,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.2897,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.44695721096232843,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.2572,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.5205254112930788,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3075,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.4946185356493118,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.2929,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.48716072005272076,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.286,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.4790877341876439,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.288,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.47264830121021645,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.2946,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5006750374266931,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3081,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.4517141193217564,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.2919,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.509066093624843,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3076,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.4857232500708578,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.2744,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.4680044100341586,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.2925,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.45684519556504416,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.2811,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.47915253952522047,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.2956,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.4821235964573415,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.2816,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.4621285646373329,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.2749,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.5024441979215891,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.2968,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.5162795697505873,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.3082,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.4628424157812997,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.2724,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.5192627475932668,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.2837,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.48180801658343153,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.2924,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.5239349516936551,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.2914,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.4680001907322504,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3022,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.46903254149542406,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.2894,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.46514299307055085,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.2778,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.46786678919628893,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3165,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.4815259163629924,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.2735,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.4575874982747357,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.2941,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.4455921675251331,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.2928,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.4971912543023757,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3067,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.44776631774903275,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.2838,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.4727923684769808,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3102,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.4633585131132337,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.2764,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.4595589816187316,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.2953,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.5010722853139445,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3114,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.46130493817081925,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.2972,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.46556293702984924,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.2802,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.48811524509467297,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.2857,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.4728002538841819,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3103,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.46372154035256435,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.2874,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.4657433417907108,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3076,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.45699674843818405,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.2799,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.4548880473872137,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.2733,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.4653316615177795,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.2915,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.4475668697593295,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.2799,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.5010598666600015,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3157,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.48634651334228796,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.2873,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.46615464283084973,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3079,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.47214661807095565,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.2859,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.48238091971831226,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.2805,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.43935175636664103,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.2846,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.46787321523554454,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.295,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.4862860557985448,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.309,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.46747747545844687,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3024,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.4829180322906678,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3102,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.453547928818171,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3007,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.4448930591795825,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3013,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.4219722594914634,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.2789,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.48698090366432295,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.2981,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.4567858448982877,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.2982,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.4757209972329756,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.2811,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.4705193989366389,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.2769,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.4814816034121053,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3287,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.4619846104988216,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.2681,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.48675057718800324,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.2916,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.5043782025502958,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.2871,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.47818172666038505,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.2833,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.47406724469356204,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.2859,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5140588797895533,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.2711,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.44970796065349794,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2843,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.5065795105384991,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.2852,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.49597374985386533,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.2935,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5309818555939366,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3213,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.4491426273568869,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.27,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.4766964829241058,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.2968,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.475467522268196,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.2972,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.4449253969235535,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.296,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.46162420516037367,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.2874,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.46480074154589224,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.2885,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.4610114856820535,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.2832,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.4473371869438646,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.2751,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.4509802264277696,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.2952,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.47784544664062706,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.2828,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.4642949063758113,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3007,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.44817223031276676,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.2922,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.4563529744011171,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.271,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.46294019558966465,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3084,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.42986233532082907,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.2788,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.45618300677462487,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.2983,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.5774287823200966,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3123,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.44487127602025583,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.2854,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.4472344857979044,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.2684,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.43303565606668126,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.2884,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.46183302826145173,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.2969,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.4424459067576534,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.2811,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.4291498049395363,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.2865,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.46653630937009316,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3118,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.48714728905370125,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.2911,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.47238243030232885,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.2877,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.4737619755246775,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.2861,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.49837952354197734,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3064,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.4602720411618602,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.2834,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.4678725930615326,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.2971,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.4344727265229032,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.2747,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.4780782617119267,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.2907,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.45487972028615903,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.2794,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.4513689656849026,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.278,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.4758272903426527,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.2998,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.4296297751598368,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.2782,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.45948160103189983,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.2892,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.46599207319265545,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3147,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.4725535465521289,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.2911,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.4202521889229179,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.2559,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 1.3361809010830705,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.4105,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.46175703252187217,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.2936,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.4480411665814551,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.2957,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.46452612552510364,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.3052,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.4465379386678439,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.2878,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.46935090944157487,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3088,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.44441241907018664,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.2877,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.43739729019876006,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.2668,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.49049938285225253,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3306,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.43949732647755024,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.2945,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.44274302895553463,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.2771,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.4509764671850016,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.2916,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.46450606697442026,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.2803,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.4399875238248579,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.2982,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.4765382646360091,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.2978,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.4311950226130485,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.2769,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.4818177275404289,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3193,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.44300611484214525,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.2768,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.4224446353596823,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.2681,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.44254828306914823,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.2877,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.4870698484838611,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.2954,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.47907288376995916,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3069,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.4484669100005068,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.2712,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.4715403262598973,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.2941,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.4406560753111995,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.2922,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.47758353600803277,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.308,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.44546257932571637,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.2721,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.4207280436832926,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.2655,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.4635694102114165,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3043,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.40656081182505616,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.2681,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.4394824677631399,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.283,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.45359102665650314,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3005,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.4598513201416267,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.2991,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.4386889560816277,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.278,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.4937800311880567,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3136,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.44730174209609136,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.288,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.4494127285360109,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.2857,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.47385401485036055,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.2941,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.46137769278204527,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.2963,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.44350749538891765,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.2732,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.48086031985728594,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.302,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.45202010900920186,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.2902,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.44275909517428036,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.277,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.45843520630804896,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.2982,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.45775299914552603,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.2938,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.4647675074930916,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.281,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.43523835897388274,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.285,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.437699491851768,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.2667,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.4813183660158604,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.2933,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.4165396925396624,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.2863,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.44518965920414766,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.2847,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.4465051544046675,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.2793,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4577693285635657,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.2932,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.4550813295854022,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.2999,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.5270361298018591,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.3145,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.45635774759700926,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.2874,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.4548210187622903,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.2916,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.41764700732210985,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.2694,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.44404331670716474,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.3078,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.4223992983493036,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.2766,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.4649732841727845,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3089,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.4345036260797423,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.2827,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.4505777506331994,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3001,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.43025433513610284,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.2851,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.4610971891812719,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3121,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.41989401745871574,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.2777,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.43587371084845594,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.2981,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.4081587389802936,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.2616,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.42612822085866214,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.2768,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.4439394782137643,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3011,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.4322724173867146,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.2797,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.4463082031241676,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.2891,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.4361898422949541,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.2796,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.4755428102223598,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3129,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.46071829540545417,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.2974,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.45140286266139257,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.2845,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.4434042041065003,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.2786,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.4391457650084251,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.2967,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.4782349557475675,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3118,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.4377570181298683,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.2796,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.4422176230321507,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.2936,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.44753412420824884,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.2806,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.47986311381683994,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.2548,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.574253734292841,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.1613,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.4531371502674556,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.143,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.3788966036338224,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.13,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.4734088416824507,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1298,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.6355957950264024,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.1456,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.5479697596146982,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1422,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.48411170115997254,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.1552,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.3923247896155167,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.133,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.37808437435502373,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1235,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.4300753300106287,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1402,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.40888867951961794,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1346,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.38926197155696934,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1322,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.36847629378939634,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.1278,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.4075846561332198,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1417,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.38788166593986023,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.1265,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.43930742732958084,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1474,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.38464968172246006,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.1279,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.4048152199840416,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1346,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.4093849269647369,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.1442,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.37013181856489524,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1265,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.3888442777184902,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.143,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.36714512543854616,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1239,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.37942579103340784,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1288,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.3619343659182596,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1215,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.41610861493188733,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1409,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.38812137274489206,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1315,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.41145113605349565,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.1365,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.3975352480396441,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1348,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.4055932346217707,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1412,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.3653608182071289,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1274,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.36762784461926934,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1218,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.4251659742603981,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.1552,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.37818809265273934,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1206,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.37936415384505845,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1301,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.4110171828370333,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1343,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.3797681478294375,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1202,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.4036766086149422,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.137,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.3959275251027369,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1359,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 1.9344241540672547,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2128,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.3874127045864584,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.1251,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.3930634123945566,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1303,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.4097428588817809,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1404,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.38603221231715373,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1282,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.37965331224785637,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.1322,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.3817513456484927,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1299,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.39802292627498365,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.1397,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.40972221573407136,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1468,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.3633715934477945,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1212,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.38083832686962704,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1308,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.3947042091633704,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.1396,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.4014064926457807,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1324,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.3890726294520334,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.1345,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.379043820175968,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1217,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.38271172143020044,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.1269,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.385330782965136,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.1294,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.41263483515867466,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.156,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.3822781831300798,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.1287,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.4073056931158842,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1375,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.37964256277456154,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.1398,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.38804928771501795,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.1372,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.3841728240023331,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1309,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.3787177316920125,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.13,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.35951613950556244,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.1213,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.426670172952506,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1449,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.3827270716393889,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1342,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.3792888343677404,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1301,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.41713352427104733,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.146,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.39286071588956445,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1418,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.38334079162000056,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1279,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.3998274277718388,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1364,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.37654596292156894,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1263,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.3998044384910083,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.1456,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.3611191829413262,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.1324,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.36615980883431487,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1328,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.3869869466114801,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.148,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.3769808369436343,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.1338,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.3729892314467185,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1261,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.36488446375436323,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1302,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.38061341052177433,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1309,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.3744162400925357,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.1339,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.3940890979169395,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.137,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.38460218002425617,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1293,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.3848794887083199,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1419,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.3836854713252871,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1294,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.40404289114729264,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.1347,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.39085360923997553,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1338,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.3911629543961157,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.1316,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.37022158259193394,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.1197,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.37406921691827066,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.1281,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.3886658639315364,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1315,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.3646686641098191,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.1221,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.40419172530319014,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1432,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.3781087206999772,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1327,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.3768417458570598,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.1303,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.39698450053123213,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1344,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.37109054582765666,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.1209,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.40086619742658713,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1451,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.37362172653493925,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1231,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.39439926468965186,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.122,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.3912173755179616,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.139,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.37998202415931404,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.133,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.37692337490602656,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1253,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.3926940758975541,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1323,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.3741377244983654,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1292,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.40808538372293424,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1388,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.3986604987581128,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1326,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.38465286709653557,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.1301,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.37000880713262996,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1239,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.37572619794657663,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.1294,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.39029175355856455,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1432,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.3685807125774216,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1224,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.4205497734557107,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1323,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.37163378322511437,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1158,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.3689649530508869,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1238,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.4069081726680708,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.1404,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.36380000040959964,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1176,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.3866393090917068,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1305,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.38475691582160015,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1355,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.37423228994507257,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.1292,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.365961306351135,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1252,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.3761190496541888,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1305,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.4016585637580695,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1354,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.36615608082959733,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.1282,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.3815503325370605,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1355,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.3986957607039675,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.1387,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.3565691341376036,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.1185,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.37518395100751206,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1261,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.4002873593553736,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.135,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.36859384713383375,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1191,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.38175404225553594,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1328,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.3736671585683549,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1227,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.3918102192261629,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1381,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.3775529985688956,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1296,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.3941510687977733,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1553,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.397248425706653,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.124,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.3690661108771123,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1214,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.3844056293677597,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1432,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.385953496222858,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.1335,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.38979040869340686,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1355,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.3564814552963281,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1304,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.3856397415726148,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1352,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.369127647463776,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.1175,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.3600952021729805,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.124,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.3860376822237948,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1357,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.38753791661969633,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1381,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.35676793537595414,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1231,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.3874664779139385,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.131,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.37890461966121564,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1234,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.3978706827461494,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1405,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.36526319751757813,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1209,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.3766979020169153,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1253,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.39213004462213064,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1248,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.3861020373269114,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1338,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.3803146769639753,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1286,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.3754804418380057,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1266,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.3810559956236822,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1311,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.39373483286452315,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1323,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.38261687136577555,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.138,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.36710681826607483,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1297,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.3504193533492048,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1153,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.394223207831258,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1416,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.3721710488937098,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.1297,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.38515757814284696,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1339,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.3879133911813841,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1299,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.3967306773616421,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1376,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.40345198971668206,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1335,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.36535955395290914,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1315,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.3792664577195249,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1375,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.3597035269368052,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1217,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.39440555198645816,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1429,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.37453821820151273,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1281,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.3531297615183325,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1146,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.38496120270991163,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1311,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.4009889466891926,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1288,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.37803554205680484,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1237,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.4001666615816794,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1275,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.3748705754106017,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1251,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.42346950078865553,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.1448,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.373354742767265,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.1231,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.3815396494899487,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1325,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.3840394376284412,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1394,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.36687124794835124,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1279,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.3922580887905504,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1336,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.36932220844015556,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.118,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.3698475079892307,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1188,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.39143813469831956,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1338,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.3725955883470649,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.124,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.3953662331223453,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1324,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.3713337893782034,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.1252,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.366980806981726,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1288,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.3800317263324656,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1296,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.37814366924160964,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1276,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.375474660683272,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1287,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.3706321879230013,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1187,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.37684364960596267,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1272,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.38129553881702694,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1354,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.35241529320927845,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1201,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.3850794235623421,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1333,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.3821538213479597,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.1381,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.3801696492280892,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.126,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.40697644407392053,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1432,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.3880675698267021,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1359,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.37886753949534163,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.1182,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.38517284963362197,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1259,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.3729271831844392,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1288,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.3892952742710937,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1292,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.357716069154611,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.123,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.3711527494097108,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1298,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.38083512807113973,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1418,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.39717844658067303,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1335,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.3607342535416173,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.1163,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.4134193710145695,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.149,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.39272914143896637,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1338,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.3784098042117597,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1299,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.37165129431157795,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1268,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.35631737297135246,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1131,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.3823203691303481,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1353,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.3872206551104743,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.135,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.3853904589459043,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.1328,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.3918301919502782,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1275,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.3481124581080518,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1039,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.4032529424476726,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1379,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.37253816881408813,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1297,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.36747390389046203,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1287,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.3808892950507798,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1225,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.3752789418036457,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1349,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.3775088117547138,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1397,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.3672187678942556,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.1204,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.37375764211312856,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1264,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.36824013155013524,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1174,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.3626771392353597,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1219,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.3917177806268977,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1311,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.38277770582189546,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1314,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.3505333387026878,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1186,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.39621660767573724,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.1453,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.35907136511897836,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1263,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.35225538409388957,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1223,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.3742955250091741,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1226,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.47756111635747533,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1529,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.3733867442492367,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1225,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.36397508882750457,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1185,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.36516340879113485,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1294,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.368952614305248,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.128,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.358603693442455,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1196,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.4407569729366144,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1492,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.3770258845440762,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.1245,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.3636708114984919,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.4113838756733195,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.0677,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.32873707533803065,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.056,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.33903668795673103,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.0671,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.28648918894593195,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.0573,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.28104119970296626,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.0659,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.2789227118201246,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.0641,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.2796711048038886,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.0661,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.29663381386967086,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.0561,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.3322736721617918,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.0594,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.35075753481932753,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.0612,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.3682863762879697,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.0632,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.3448938679851067,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.0596,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.32299532206803017,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.0559,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.3471560336861636,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.0648,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.2836099445005014,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.0531,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.3006770798650389,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.0587,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.28544504827064665,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.0573,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.28710029027295053,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.0583,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.2727525178334994,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.0588,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.2711809256157805,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.0586,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.2603458817080641,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.0573,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.2932374544320147,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.0705,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.2667476478503739,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.0571,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.27672847156928826,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.0592,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.2596338301627828,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.0556,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.2876680492773809,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.0649,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.2672567419021313,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.0546,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.2743617885813735,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.0584,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.269952533097825,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.0557,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.28995746086202084,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.0649,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.2637162544421887,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.0529,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.27518515910336433,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.0551,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.27578195163830443,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.06,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.29231494949581593,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.0607,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.27350348664840257,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.0552,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.279919405011891,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.0603,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.272655595329232,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.0549,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.2878454401688556,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.0591,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.28483453037449286,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.0653,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.289898729173417,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.0587,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.25490786896178685,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.0505,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.284969641499039,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.0597,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.2918056571234901,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.065,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.2725576218981918,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.0568,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.2949987329280441,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.0648,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.29173210155471924,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.0621,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.2629541405535745,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.0539,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.27302329841868556,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.0566,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.2810076433529102,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.0616,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.26797546849202797,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.0608,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.2826237441290358,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.0578,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.27011148889146214,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.0538,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.28601734094945247,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.062,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.2836158517352792,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.0596,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.27765713787859564,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.0594,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.27627366706994555,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.0545,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.2596189970565373,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.048,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.31223178732992135,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.0822,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.2642963484772625,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.0522,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.27456763987370897,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.056,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.2680592949120706,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.0604,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.2727955253406336,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.0554,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.2790723960765753,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.0554,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.2823458168505066,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.0586,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.28696377223625846,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.0628,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.26110653501356557,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.0505,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.27845774853879046,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.0599,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.2775611839470655,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.0563,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.281081461479417,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.0581,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.27863977588321115,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.0623,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.25850084104301774,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.0544,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.28467136355083744,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.0577,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.2702957297111791,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.0531,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.28791279105840045,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.0593,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.27564414572316265,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.0554,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.2829681318209881,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.055,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.26354499702996587,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.0525,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.28549045766824915,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.0631,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.265842469605766,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.0509,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.2826219576076134,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.061,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.2725770750551493,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.0545,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.28312428666665457,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.0602,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.27593590699208714,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.0595,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.26507256310262695,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.0527,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.2853620734470677,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.0649,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.24859255585728232,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.0502,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.29414487451923244,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.0634,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.2620585171752594,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.0536,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.3044258068441641,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.06,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.26956487677230934,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.0551,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.28330511521725527,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.0623,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.2667023997074734,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.0519,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.27442875193062743,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.0622,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.2739785587402866,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.0569,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.26150664648151223,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.0549,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.27035415883510483,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.0596,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.2559711266495521,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.0538,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.2782360331945333,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.0569,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.2956136727619104,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.0593,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.2701461944478313,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.058,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.2679460294452299,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.0552,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 4.149923783173249,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1699,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.27945099608387697,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.0543,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.27270917617559587,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.0555,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.2781298471132881,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.0615,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.2773187704820607,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.0582,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.28387139733122563,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.0604,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.268430548346151,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.0548,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.27669222259422344,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.0564,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.2682057327847503,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.0518,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.2735674177834583,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.0576,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.278103095572561,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.0547,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.27150274806179203,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.0561,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.26664024808450865,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.0533,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.2764847341902885,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.0585,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.28641607755879084,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.0586,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.2734341816072021,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.0552,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.27188947127350727,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.0534,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.28326036366175317,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.0608,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.27182349086183344,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.0527,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.2851854479036507,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.0626,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.27550564811018174,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.0633,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.2572258263123181,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.0496,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.28650285458629954,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.0599,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.27113130470294067,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.056,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.2688982729778506,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.054,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.30088089964209586,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.0646,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.30113775843234253,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.0671,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.2587343279651594,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.053,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.27507766816177326,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.0575,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.27825682340578706,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.0572,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.272886268365024,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.0553,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.27513575238236904,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.0555,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.27425596924018497,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.0535,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.2822739056208061,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.0548,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.277813014288407,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.057,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.26975493359943387,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.0542,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.2599025832969597,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.0516,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.2861715848682678,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.0577,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.2791862262108885,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.0578,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.2810743036905506,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.0613,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.2702376810553738,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.0543,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.2926125741161434,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.0641,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.2961129898959361,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.0637,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.2753631773934487,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.0575,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.26131139787884855,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.0564,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.2797418893156601,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.0596,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.2770768671316833,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.0595,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.2718676923176175,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.0539,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.30600108121503705,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.0668,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.26176887189024306,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.0516,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.2613829120485367,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.0518,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.27752054830432343,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.0562,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.26738501961985484,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.054,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.26044265313374443,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.0531,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.26973848956264207,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.0558,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.27322893144052973,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.0544,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.2829721933451359,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.0632,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.2900985533730961,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.0616,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.2698135635390665,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.0534,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.27623412420834437,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.0578,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.28741536607574314,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.0629,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.2858977663010505,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.0577,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.2727807924599147,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.0545,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.2671487906212562,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.0531,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.25653403872696756,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.0496,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.27916894536396286,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.0579,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.2860199222292299,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.0617,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.26257025385591637,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.0514,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.28072885842433537,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.059,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.2615161118538124,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.0539,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.26032149069948995,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.0511,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.27562851712612635,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.0565,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.2653534765230234,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.0546,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.2736739110989586,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.0532,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.28390883677539686,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.0578,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.27609528055325294,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.0603,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.278651791032142,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.061,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.26392231050396564,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.0511,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.28420328676659407,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.059,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.29107235852480723,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.06,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.2649138876238062,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.053,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.2675077404598596,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.0514,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.2789633285339604,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.0572,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.26781509076161986,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.0564,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.2610300883686458,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.0495,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.2896258025628451,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.0617,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.2659167539489299,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.0505,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.2807997486710564,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.0628,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.27736555308486144,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.0564,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.26862250119096737,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.053,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.27330002518858676,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.0543,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.27117994135840034,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.0558,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.2645291763643275,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.0548,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.2716834766801666,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.0567,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.26886429590115074,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.0524,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.3069876610452829,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.0656,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.26458613405593134,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.0525,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.2943019582892251,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.0602,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.2653020634987694,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.0494,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.2718441571312757,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.054,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.28618913728927103,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.0628,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.2652571596665405,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.0523,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.33461084093968607,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.0694,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.2849256988741649,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.066,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.2635099340569179,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.0562,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.26987643051006405,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.0557,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.2699077686610517,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.0554,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.2673003290669875,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.0528,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.30756102958818654,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.0735,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.2608197253414941,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.0508,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.2823578922373537,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.0602,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.28600903853784565,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.0597,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.2693341453991124,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.0528,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.27949842000589503,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.0583,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.27949322165466595,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.0558,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.27548473296327836,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.0579,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.26476178258331023,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.052,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.2849541859808429,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.0596,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.2614188321180138,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.0556,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.26427557834532545,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.05,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.2822407726703744,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.0597,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.24564378490305047,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.0484,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.2741489682907927,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.0584,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.2706829909246734,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.0569,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.2839750532856087,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.0621,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.2618274650343186,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.054,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.2876299686455689,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.0576,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.281991996507193,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.0606,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.242865933189303,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.0486,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.2684696869219734,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.0572,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.2873236395584926,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.0563,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.2875612088355685,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.0572,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.26146154402663774,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.0517,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.27661211569561667,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.0597,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.27169619344620616,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.0555,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.27173333205167577,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.0559,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.26353257673619557,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.0539,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.26583597894361133,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.0547,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.25397978708418284,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.0559,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.2933831199444266,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.0605,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.3192523400458108,
+      "learning_rate": 0.0,
+      "loss": 0.0673,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 1.2680760044552192e+18,
+      "train_loss": 0.34186193917582675,
+      "train_runtime": 54145.9256,
+      "train_samples_per_second": 2.918,
+      "train_steps_per_second": 0.023
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2680760044552192e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}