{
  "best_metric": 6.0493364334106445,
  "best_model_checkpoint": "HBERTv1_emb_compress_48_L10_H256_A4/checkpoint-450000",
  "epoch": 5.0,
  "eval_steps": 10000,
  "global_step": 457720,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 5.000000000000001e-07,
      "loss": 10.2451,
      "step": 500
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 10.1165,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.5e-06,
      "loss": 10.0395,
      "step": 1500
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 9.9311,
      "step": 2000
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.5e-06,
      "loss": 9.7921,
      "step": 2500
    },
    {
      "epoch": 0.03,
      "learning_rate": 3e-06,
      "loss": 9.6257,
      "step": 3000
    },
    {
      "epoch": 0.04,
      "learning_rate": 3.5e-06,
      "loss": 9.4358,
      "step": 3500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.000000000000001e-06,
      "loss": 9.2227,
      "step": 4000
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.5e-06,
      "loss": 8.9936,
      "step": 4500
    },
    {
      "epoch": 0.05,
      "learning_rate": 5e-06,
      "loss": 8.7508,
      "step": 5000
    },
    {
      "epoch": 0.06,
      "learning_rate": 5.500000000000001e-06,
      "loss": 8.5027,
      "step": 5500
    },
    {
      "epoch": 0.07,
      "learning_rate": 6e-06,
      "loss": 8.2533,
      "step": 6000
    },
    {
      "epoch": 0.07,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 8.0151,
      "step": 6500
    },
    {
      "epoch": 0.08,
      "learning_rate": 7e-06,
      "loss": 7.793,
      "step": 7000
    },
    {
      "epoch": 0.08,
      "learning_rate": 7.500000000000001e-06,
      "loss": 7.5987,
      "step": 7500
    },
    {
      "epoch": 0.09,
      "learning_rate": 8.000000000000001e-06,
      "loss": 7.4367,
      "step": 8000
    },
    {
      "epoch": 0.09,
      "learning_rate": 8.5e-06,
      "loss": 7.3083,
      "step": 8500
    },
    {
      "epoch": 0.1,
      "learning_rate": 9e-06,
      "loss": 7.2285,
      "step": 9000
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.5e-06,
      "loss": 7.1671,
      "step": 9500
    },
    {
      "epoch": 0.11,
      "learning_rate": 1e-05,
      "loss": 7.1164,
      "step": 10000
    },
    {
      "epoch": 0.11,
      "eval_accuracy": 0.08299088710181593,
      "eval_loss": 7.096699237823486,
      "eval_runtime": 764.9467,
      "eval_samples_per_second": 403.108,
      "eval_steps_per_second": 6.3,
      "step": 10000
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.988832305905477e-06,
      "loss": 7.0743,
      "step": 10500
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.977664611810955e-06,
      "loss": 7.0306,
      "step": 11000
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.966496917716431e-06,
      "loss": 6.9993,
      "step": 11500
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.955329223621908e-06,
      "loss": 6.9675,
      "step": 12000
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.944161529527384e-06,
      "loss": 6.9456,
      "step": 12500
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.93299383543286e-06,
      "loss": 6.9163,
      "step": 13000
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.921826141338338e-06,
      "loss": 6.895,
      "step": 13500
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.910658447243815e-06,
      "loss": 6.8654,
      "step": 14000
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.899490753149291e-06,
      "loss": 6.8518,
      "step": 14500
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.888323059054767e-06,
      "loss": 6.8237,
      "step": 15000
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.877155364960243e-06,
      "loss": 6.8098,
      "step": 15500
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.865987670865721e-06,
      "loss": 6.793,
      "step": 16000
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.854819976771198e-06,
      "loss": 6.7751,
      "step": 16500
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.843652282676674e-06,
      "loss": 6.7629,
      "step": 17000
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.83248458858215e-06,
      "loss": 6.7506,
      "step": 17500
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.821316894487627e-06,
      "loss": 6.7416,
      "step": 18000
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.810149200393103e-06,
      "loss": 6.7204,
      "step": 18500
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.798981506298581e-06,
      "loss": 6.7151,
      "step": 19000
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.787813812204057e-06,
      "loss": 6.7109,
      "step": 19500
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.776646118109534e-06,
      "loss": 6.694,
      "step": 20000
    },
    {
      "epoch": 0.22,
      "eval_accuracy": 0.10645789734965333,
      "eval_loss": 6.6867475509643555,
      "eval_runtime": 763.7287,
      "eval_samples_per_second": 403.751,
      "eval_steps_per_second": 6.31,
      "step": 20000
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.76547842401501e-06,
      "loss": 6.6776,
      "step": 20500
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.754310729920486e-06,
      "loss": 6.6717,
      "step": 21000
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.743143035825964e-06,
      "loss": 6.6642,
      "step": 21500
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.73197534173144e-06,
      "loss": 6.6555,
      "step": 22000
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.720807647636917e-06,
      "loss": 6.6516,
      "step": 22500
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.709639953542393e-06,
      "loss": 6.6398,
      "step": 23000
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.69847225944787e-06,
      "loss": 6.6302,
      "step": 23500
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.687304565353347e-06,
      "loss": 6.6274,
      "step": 24000
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.676136871258824e-06,
      "loss": 6.6209,
      "step": 24500
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.6649691771643e-06,
      "loss": 6.608,
      "step": 25000
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.653801483069776e-06,
      "loss": 6.602,
      "step": 25500
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.642633788975253e-06,
      "loss": 6.5966,
      "step": 26000
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.63146609488073e-06,
      "loss": 6.5982,
      "step": 26500
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.620298400786207e-06,
      "loss": 6.5855,
      "step": 27000
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.609130706691683e-06,
      "loss": 6.57,
      "step": 27500
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.59796301259716e-06,
      "loss": 6.5761,
      "step": 28000
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.586795318502636e-06,
      "loss": 6.5617,
      "step": 28500
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.575627624408112e-06,
      "loss": 6.5633,
      "step": 29000
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.56445993031359e-06,
      "loss": 6.5537,
      "step": 29500
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.553292236219066e-06,
      "loss": 6.545,
      "step": 30000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.1171236652942363,
      "eval_loss": 6.54450798034668,
      "eval_runtime": 765.3292,
      "eval_samples_per_second": 402.906,
      "eval_steps_per_second": 6.297,
      "step": 30000
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.542124542124543e-06,
      "loss": 6.543,
      "step": 30500
    },
    {
      "epoch": 0.34,
      "learning_rate": 9.530956848030019e-06,
      "loss": 6.5448,
      "step": 31000
    },
    {
      "epoch": 0.34,
      "learning_rate": 9.519789153935495e-06,
      "loss": 6.531,
      "step": 31500
    },
    {
      "epoch": 0.35,
      "learning_rate": 9.508621459840973e-06,
      "loss": 6.5307,
      "step": 32000
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.49745376574645e-06,
      "loss": 6.5241,
      "step": 32500
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.486286071651926e-06,
      "loss": 6.5179,
      "step": 33000
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.475118377557404e-06,
      "loss": 6.5095,
      "step": 33500
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.463950683462878e-06,
      "loss": 6.5053,
      "step": 34000
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.452782989368356e-06,
      "loss": 6.4983,
      "step": 34500
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.441615295273833e-06,
      "loss": 6.4981,
      "step": 35000
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.430447601179309e-06,
      "loss": 6.4982,
      "step": 35500
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.419279907084787e-06,
      "loss": 6.4951,
      "step": 36000
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.408112212990263e-06,
      "loss": 6.481,
      "step": 36500
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.396944518895738e-06,
      "loss": 6.4846,
      "step": 37000
    },
    {
      "epoch": 0.41,
      "learning_rate": 9.385776824801216e-06,
      "loss": 6.4752,
      "step": 37500
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.374609130706692e-06,
      "loss": 6.4735,
      "step": 38000
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.36344143661217e-06,
      "loss": 6.4718,
      "step": 38500
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.352273742517646e-06,
      "loss": 6.4597,
      "step": 39000
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.341106048423123e-06,
      "loss": 6.4541,
      "step": 39500
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.329938354328599e-06,
      "loss": 6.4556,
      "step": 40000
    },
    {
      "epoch": 0.44,
      "eval_accuracy": 0.12498217135629651,
      "eval_loss": 6.452664375305176,
      "eval_runtime": 765.1792,
      "eval_samples_per_second": 402.985,
      "eval_steps_per_second": 6.298,
      "step": 40000
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.318770660234075e-06,
      "loss": 6.4531,
      "step": 40500
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.307602966139552e-06,
      "loss": 6.4469,
      "step": 41000
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.29643527204503e-06,
      "loss": 6.4433,
      "step": 41500
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.285267577950506e-06,
      "loss": 6.4412,
      "step": 42000
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.274099883855982e-06,
      "loss": 6.4407,
      "step": 42500
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.262932189761459e-06,
      "loss": 6.4402,
      "step": 43000
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.251764495666935e-06,
      "loss": 6.4222,
      "step": 43500
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.240596801572413e-06,
      "loss": 6.4262,
      "step": 44000
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.229429107477889e-06,
      "loss": 6.42,
      "step": 44500
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.218261413383365e-06,
      "loss": 6.4221,
      "step": 45000
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.207093719288842e-06,
      "loss": 6.4172,
      "step": 45500
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.195926025194318e-06,
      "loss": 6.4142,
      "step": 46000
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.184758331099796e-06,
      "loss": 6.4107,
      "step": 46500
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.173590637005272e-06,
      "loss": 6.4031,
      "step": 47000
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.162422942910749e-06,
      "loss": 6.4031,
      "step": 47500
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.151255248816225e-06,
      "loss": 6.3971,
      "step": 48000
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.140087554721701e-06,
      "loss": 6.3941,
      "step": 48500
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.12891986062718e-06,
      "loss": 6.3968,
      "step": 49000
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.117752166532655e-06,
      "loss": 6.3975,
      "step": 49500
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.106584472438132e-06,
      "loss": 6.3891,
      "step": 50000
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.1305474526564509,
      "eval_loss": 6.3831257820129395,
      "eval_runtime": 765.2762,
      "eval_samples_per_second": 402.934,
      "eval_steps_per_second": 6.297,
      "step": 50000
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.095416778343608e-06,
      "loss": 6.3872,
      "step": 50500
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.084249084249084e-06,
      "loss": 6.3793,
      "step": 51000
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.07308139015456e-06,
      "loss": 6.3784,
      "step": 51500
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.061913696060039e-06,
      "loss": 6.3763,
      "step": 52000
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.050746001965515e-06,
      "loss": 6.37,
      "step": 52500
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.039578307870991e-06,
      "loss": 6.3721,
      "step": 53000
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.028410613776468e-06,
      "loss": 6.3695,
      "step": 53500
    },
    {
      "epoch": 0.59,
      "learning_rate": 9.017242919681944e-06,
      "loss": 6.3702,
      "step": 54000
    },
    {
      "epoch": 0.6,
      "learning_rate": 9.006075225587422e-06,
      "loss": 6.3597,
      "step": 54500
    },
    {
      "epoch": 0.6,
      "learning_rate": 8.994907531492898e-06,
      "loss": 6.3581,
      "step": 55000
    },
    {
      "epoch": 0.61,
      "learning_rate": 8.983739837398374e-06,
      "loss": 6.3604,
      "step": 55500
    },
    {
      "epoch": 0.61,
      "learning_rate": 8.97257214330385e-06,
      "loss": 6.3592,
      "step": 56000
    },
    {
      "epoch": 0.62,
      "learning_rate": 8.961404449209327e-06,
      "loss": 6.3494,
      "step": 56500
    },
    {
      "epoch": 0.62,
      "learning_rate": 8.950236755114805e-06,
      "loss": 6.3539,
      "step": 57000
    },
    {
      "epoch": 0.63,
      "learning_rate": 8.939069061020281e-06,
      "loss": 6.3484,
      "step": 57500
    },
    {
      "epoch": 0.63,
      "learning_rate": 8.927901366925758e-06,
      "loss": 6.3429,
      "step": 58000
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.916733672831236e-06,
      "loss": 6.3408,
      "step": 58500
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.90556597873671e-06,
      "loss": 6.3404,
      "step": 59000
    },
    {
      "epoch": 0.65,
      "learning_rate": 8.894398284642188e-06,
      "loss": 6.333,
      "step": 59500
    },
    {
      "epoch": 0.66,
      "learning_rate": 8.883230590547665e-06,
      "loss": 6.3404,
      "step": 60000
    },
    {
      "epoch": 0.66,
      "eval_accuracy": 0.1350277249660963,
      "eval_loss": 6.333357810974121,
      "eval_runtime": 765.1505,
      "eval_samples_per_second": 403.0,
      "eval_steps_per_second": 6.298,
      "step": 60000
    },
    {
      "epoch": 0.66,
      "learning_rate": 8.87206289645314e-06,
      "loss": 6.336,
      "step": 60500
    },
    {
      "epoch": 0.67,
      "learning_rate": 8.860895202358619e-06,
      "loss": 6.3311,
      "step": 61000
    },
    {
      "epoch": 0.67,
      "learning_rate": 8.849727508264095e-06,
      "loss": 6.3318,
      "step": 61500
    },
    {
      "epoch": 0.68,
      "learning_rate": 8.83855981416957e-06,
      "loss": 6.3274,
      "step": 62000
    },
    {
      "epoch": 0.68,
      "learning_rate": 8.827392120075048e-06,
      "loss": 6.3242,
      "step": 62500
    },
    {
      "epoch": 0.69,
      "learning_rate": 8.816224425980524e-06,
      "loss": 6.3212,
      "step": 63000
    },
    {
      "epoch": 0.69,
      "learning_rate": 8.805056731886002e-06,
      "loss": 6.3174,
      "step": 63500
    },
    {
      "epoch": 0.7,
      "learning_rate": 8.793889037791478e-06,
      "loss": 6.3167,
      "step": 64000
    },
    {
      "epoch": 0.7,
      "learning_rate": 8.782721343696955e-06,
      "loss": 6.3179,
      "step": 64500
    },
    {
      "epoch": 0.71,
      "learning_rate": 8.771553649602431e-06,
      "loss": 6.3098,
      "step": 65000
    },
    {
      "epoch": 0.72,
      "learning_rate": 8.760385955507907e-06,
      "loss": 6.3104,
      "step": 65500
    },
    {
      "epoch": 0.72,
      "learning_rate": 8.749218261413384e-06,
      "loss": 6.3088,
      "step": 66000
    },
    {
      "epoch": 0.73,
      "learning_rate": 8.738050567318861e-06,
      "loss": 6.3119,
      "step": 66500
    },
    {
      "epoch": 0.73,
      "learning_rate": 8.726882873224338e-06,
      "loss": 6.3092,
      "step": 67000
    },
    {
      "epoch": 0.74,
      "learning_rate": 8.715715179129814e-06,
      "loss": 6.3048,
      "step": 67500
    },
    {
      "epoch": 0.74,
      "learning_rate": 8.70454748503529e-06,
      "loss": 6.3017,
      "step": 68000
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.693379790940767e-06,
      "loss": 6.3028,
      "step": 68500
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.682212096846245e-06,
      "loss": 6.2981,
      "step": 69000
    },
    {
      "epoch": 0.76,
      "learning_rate": 8.671044402751721e-06,
      "loss": 6.2955,
      "step": 69500
    },
    {
      "epoch": 0.76,
      "learning_rate": 8.659876708657197e-06,
      "loss": 6.2962,
      "step": 70000
    },
    {
      "epoch": 0.76,
      "eval_accuracy": 0.13768265237919167,
      "eval_loss": 6.294034481048584,
      "eval_runtime": 767.1588,
      "eval_samples_per_second": 401.945,
      "eval_steps_per_second": 6.282,
      "step": 70000
    },
    {
      "epoch": 0.77,
      "learning_rate": 8.648709014562674e-06,
      "loss": 6.2946,
      "step": 70500
    },
    {
      "epoch": 0.78,
      "learning_rate": 8.63754132046815e-06,
      "loss": 6.2856,
      "step": 71000
    },
    {
      "epoch": 0.78,
      "learning_rate": 8.626373626373628e-06,
      "loss": 6.2871,
      "step": 71500
    },
    {
      "epoch": 0.79,
      "learning_rate": 8.615205932279104e-06,
      "loss": 6.2857,
      "step": 72000
    },
    {
      "epoch": 0.79,
      "learning_rate": 8.60403823818458e-06,
      "loss": 6.2828,
      "step": 72500
    },
    {
      "epoch": 0.8,
      "learning_rate": 8.592870544090057e-06,
      "loss": 6.2839,
      "step": 73000
    },
    {
      "epoch": 0.8,
      "learning_rate": 8.581702849995533e-06,
      "loss": 6.2789,
      "step": 73500
    },
    {
      "epoch": 0.81,
      "learning_rate": 8.57053515590101e-06,
      "loss": 6.2883,
      "step": 74000
    },
    {
      "epoch": 0.81,
      "learning_rate": 8.559367461806487e-06,
      "loss": 6.281,
      "step": 74500
    },
    {
      "epoch": 0.82,
      "learning_rate": 8.548199767711964e-06,
      "loss": 6.2829,
      "step": 75000
    },
    {
      "epoch": 0.82,
      "learning_rate": 8.53703207361744e-06,
      "loss": 6.2733,
      "step": 75500
    },
    {
      "epoch": 0.83,
      "learning_rate": 8.525864379522916e-06,
      "loss": 6.2745,
      "step": 76000
    },
    {
      "epoch": 0.84,
      "learning_rate": 8.514696685428393e-06,
      "loss": 6.2694,
      "step": 76500
    },
    {
      "epoch": 0.84,
      "learning_rate": 8.50352899133387e-06,
      "loss": 6.2748,
      "step": 77000
    },
    {
      "epoch": 0.85,
      "learning_rate": 8.492361297239347e-06,
      "loss": 6.2735,
      "step": 77500
    },
    {
      "epoch": 0.85,
      "learning_rate": 8.481193603144823e-06,
      "loss": 6.2674,
      "step": 78000
    },
    {
      "epoch": 0.86,
      "learning_rate": 8.4700259090503e-06,
      "loss": 6.2662,
      "step": 78500
    },
    {
      "epoch": 0.86,
      "learning_rate": 8.458858214955776e-06,
      "loss": 6.2595,
      "step": 79000
    },
    {
      "epoch": 0.87,
      "learning_rate": 8.447690520861254e-06,
      "loss": 6.2599,
      "step": 79500
    },
    {
      "epoch": 0.87,
      "learning_rate": 8.43652282676673e-06,
      "loss": 6.2669,
      "step": 80000
    },
    {
      "epoch": 0.87,
      "eval_accuracy": 0.13975203188129762,
      "eval_loss": 6.262859344482422,
      "eval_runtime": 766.0897,
      "eval_samples_per_second": 402.506,
      "eval_steps_per_second": 6.29,
      "step": 80000
    },
    {
      "epoch": 0.88,
      "learning_rate": 8.425355132672206e-06,
      "loss": 6.2692,
      "step": 80500
    },
    {
      "epoch": 0.88,
      "learning_rate": 8.414187438577683e-06,
      "loss": 6.2568,
      "step": 81000
    },
    {
      "epoch": 0.89,
      "learning_rate": 8.403019744483159e-06,
      "loss": 6.258,
      "step": 81500
    },
    {
      "epoch": 0.9,
      "learning_rate": 8.391852050388637e-06,
      "loss": 6.2558,
      "step": 82000
    },
    {
      "epoch": 0.9,
      "learning_rate": 8.380684356294113e-06,
      "loss": 6.2583,
      "step": 82500
    },
    {
      "epoch": 0.91,
      "learning_rate": 8.36951666219959e-06,
      "loss": 6.2532,
      "step": 83000
    },
    {
      "epoch": 0.91,
      "learning_rate": 8.358348968105068e-06,
      "loss": 6.2537,
      "step": 83500
    },
    {
      "epoch": 0.92,
      "learning_rate": 8.347181274010542e-06,
      "loss": 6.2555,
      "step": 84000
    },
    {
      "epoch": 0.92,
      "learning_rate": 8.336013579916018e-06,
      "loss": 6.2484,
      "step": 84500
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.324845885821496e-06,
      "loss": 6.2545,
      "step": 85000
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.313678191726973e-06,
      "loss": 6.2524,
      "step": 85500
    },
    {
      "epoch": 0.94,
      "learning_rate": 8.30251049763245e-06,
      "loss": 6.2457,
      "step": 86000
    },
    {
      "epoch": 0.94,
      "learning_rate": 8.291342803537927e-06,
      "loss": 6.2493,
      "step": 86500
    },
    {
      "epoch": 0.95,
      "learning_rate": 8.280175109443402e-06,
      "loss": 6.2487,
      "step": 87000
    },
    {
      "epoch": 0.96,
      "learning_rate": 8.26900741534888e-06,
      "loss": 6.2474,
      "step": 87500
    },
    {
      "epoch": 0.96,
      "learning_rate": 8.257839721254356e-06,
      "loss": 6.2424,
      "step": 88000
    },
    {
      "epoch": 0.97,
      "learning_rate": 8.246672027159832e-06,
      "loss": 6.2413,
      "step": 88500
    },
    {
      "epoch": 0.97,
      "learning_rate": 8.23550433306531e-06,
      "loss": 6.2413,
      "step": 89000
    },
    {
      "epoch": 0.98,
      "learning_rate": 8.224336638970786e-06,
      "loss": 6.2429,
      "step": 89500
    },
    {
      "epoch": 0.98,
      "learning_rate": 8.213168944876263e-06,
      "loss": 6.2352,
      "step": 90000
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.14123526312506032,
      "eval_loss": 6.236103534698486,
      "eval_runtime": 764.3112,
      "eval_samples_per_second": 403.443,
      "eval_steps_per_second": 6.305,
      "step": 90000
    },
    {
      "epoch": 0.99,
      "learning_rate": 8.202001250781739e-06,
      "loss": 6.2374,
      "step": 90500
    },
    {
      "epoch": 0.99,
      "learning_rate": 8.190833556687215e-06,
      "loss": 6.2354,
      "step": 91000
    },
    {
      "epoch": 1.0,
      "learning_rate": 8.179665862592693e-06,
      "loss": 6.2382,
      "step": 91500
    },
    {
      "epoch": 1.0,
      "learning_rate": 8.16849816849817e-06,
      "loss": 6.2349,
      "step": 92000
    },
    {
      "epoch": 1.01,
      "learning_rate": 8.157330474403646e-06,
      "loss": 6.2345,
      "step": 92500
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.146162780309122e-06,
      "loss": 6.2276,
      "step": 93000
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.134995086214599e-06,
      "loss": 6.23,
      "step": 93500
    },
    {
      "epoch": 1.03,
      "learning_rate": 8.123827392120077e-06,
      "loss": 6.2239,
      "step": 94000
    },
    {
      "epoch": 1.03,
      "learning_rate": 8.112659698025553e-06,
      "loss": 6.2275,
      "step": 94500
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.101492003931029e-06,
      "loss": 6.2273,
      "step": 95000
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.090324309836505e-06,
      "loss": 6.2227,
      "step": 95500
    },
    {
      "epoch": 1.05,
      "learning_rate": 8.079156615741982e-06,
      "loss": 6.2186,
      "step": 96000
    },
    {
      "epoch": 1.05,
      "learning_rate": 8.06798892164746e-06,
      "loss": 6.2227,
      "step": 96500
    },
    {
      "epoch": 1.06,
      "learning_rate": 8.056821227552936e-06,
      "loss": 6.2215,
      "step": 97000
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.045653533458412e-06,
      "loss": 6.2219,
      "step": 97500
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.034485839363889e-06,
      "loss": 6.2155,
      "step": 98000
    },
    {
      "epoch": 1.08,
      "learning_rate": 8.023318145269365e-06,
      "loss": 6.2237,
      "step": 98500
    },
    {
      "epoch": 1.08,
      "learning_rate": 8.012150451174841e-06,
      "loss": 6.2174,
      "step": 99000
    },
    {
      "epoch": 1.09,
      "learning_rate": 8.00098275708032e-06,
      "loss": 6.2226,
      "step": 99500
    },
    {
      "epoch": 1.09,
      "learning_rate": 7.989815062985796e-06,
      "loss": 6.2179,
      "step": 100000
    },
    {
      "epoch": 1.09,
      "eval_accuracy": 0.14288797197812605,
      "eval_loss": 6.214980602264404,
      "eval_runtime": 766.4311,
      "eval_samples_per_second": 402.327,
      "eval_steps_per_second": 6.288,
      "step": 100000
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.978647368891272e-06,
      "loss": 6.2199,
      "step": 100500
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.967479674796748e-06,
      "loss": 6.2133,
      "step": 101000
    },
    {
      "epoch": 1.11,
      "learning_rate": 7.956311980702224e-06,
      "loss": 6.2119,
      "step": 101500
    },
    {
      "epoch": 1.11,
      "learning_rate": 7.945144286607702e-06,
      "loss": 6.2146,
      "step": 102000
    },
    {
      "epoch": 1.12,
      "learning_rate": 7.933976592513179e-06,
      "loss": 6.2169,
      "step": 102500
    },
    {
      "epoch": 1.13,
      "learning_rate": 7.922808898418655e-06,
      "loss": 6.2103,
      "step": 103000
    },
    {
      "epoch": 1.13,
      "learning_rate": 7.911641204324131e-06,
      "loss": 6.2055,
      "step": 103500
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.900473510229608e-06,
      "loss": 6.202,
      "step": 104000
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.889305816135086e-06,
      "loss": 6.2099,
      "step": 104500
    },
    {
      "epoch": 1.15,
      "learning_rate": 7.878138122040562e-06,
      "loss": 6.2098,
      "step": 105000
    },
    {
      "epoch": 1.15,
      "learning_rate": 7.866970427946038e-06,
      "loss": 6.2028,
      "step": 105500
    },
    {
      "epoch": 1.16,
      "learning_rate": 7.855802733851514e-06,
      "loss": 6.1973,
      "step": 106000
    },
    {
      "epoch": 1.16,
      "learning_rate": 7.84463503975699e-06,
      "loss": 6.2036,
      "step": 106500
    },
    {
      "epoch": 1.17,
      "learning_rate": 7.833467345662469e-06,
      "loss": 6.2031,
      "step": 107000
    },
    {
      "epoch": 1.17,
      "learning_rate": 7.822299651567945e-06,
      "loss": 6.2028,
      "step": 107500
    },
    {
      "epoch": 1.18,
      "learning_rate": 7.811131957473421e-06,
      "loss": 6.2011,
      "step": 108000
    },
    {
      "epoch": 1.19,
      "learning_rate": 7.7999642633789e-06,
      "loss": 6.2014,
      "step": 108500
    },
    {
      "epoch": 1.19,
      "learning_rate": 7.788796569284374e-06,
      "loss": 6.2046,
      "step": 109000
    },
    {
      "epoch": 1.2,
      "learning_rate": 7.77762887518985e-06,
      "loss": 6.1927,
      "step": 109500
    },
    {
      "epoch": 1.2,
      "learning_rate": 7.766461181095328e-06,
      "loss": 6.191,
      "step": 110000
    },
    {
      "epoch": 1.2,
      "eval_accuracy": 0.1442675297887817,
      "eval_loss": 6.196980953216553,
      "eval_runtime": 766.3952,
      "eval_samples_per_second": 402.346,
      "eval_steps_per_second": 6.288,
      "step": 110000
    },
    {
      "epoch": 1.21,
      "learning_rate": 7.755293487000805e-06,
      "loss": 6.199,
      "step": 110500
    },
    {
      "epoch": 1.21,
      "learning_rate": 7.744125792906283e-06,
      "loss": 6.1974,
      "step": 111000
    },
    {
      "epoch": 1.22,
      "learning_rate": 7.732958098811759e-06,
      "loss": 6.2025,
      "step": 111500
    },
    {
      "epoch": 1.22,
      "learning_rate": 7.721790404717233e-06,
      "loss": 6.1973,
      "step": 112000
    },
    {
      "epoch": 1.23,
      "learning_rate": 7.710622710622711e-06,
      "loss": 6.19,
      "step": 112500
    },
    {
      "epoch": 1.23,
      "learning_rate": 7.699455016528188e-06,
      "loss": 6.1947,
      "step": 113000
    },
    {
      "epoch": 1.24,
      "learning_rate": 7.688287322433664e-06,
      "loss": 6.1912,
      "step": 113500
    },
    {
      "epoch": 1.25,
      "learning_rate": 7.677119628339142e-06,
      "loss": 6.1907,
      "step": 114000
    },
    {
      "epoch": 1.25,
      "learning_rate": 7.665951934244618e-06,
      "loss": 6.192,
      "step": 114500
    },
    {
      "epoch": 1.26,
      "learning_rate": 7.654784240150095e-06,
      "loss": 6.1948,
      "step": 115000
    },
    {
      "epoch": 1.26,
      "learning_rate": 7.643616546055571e-06,
      "loss": 6.1931,
      "step": 115500
    },
    {
      "epoch": 1.27,
      "learning_rate": 7.632448851961047e-06,
      "loss": 6.1896,
      "step": 116000
    },
    {
      "epoch": 1.27,
      "learning_rate": 7.621281157866524e-06,
      "loss": 6.1887,
      "step": 116500
    },
    {
      "epoch": 1.28,
      "learning_rate": 7.610113463772001e-06,
      "loss": 6.1882,
      "step": 117000
    },
    {
      "epoch": 1.28,
      "learning_rate": 7.598945769677477e-06,
      "loss": 6.1862,
      "step": 117500
    },
    {
      "epoch": 1.29,
      "learning_rate": 7.587778075582954e-06,
      "loss": 6.1802,
      "step": 118000
    },
    {
      "epoch": 1.29,
      "learning_rate": 7.57661038148843e-06,
      "loss": 6.1821,
      "step": 118500
    },
    {
      "epoch": 1.3,
      "learning_rate": 7.5654426873939076e-06,
      "loss": 6.182,
      "step": 119000
    },
    {
      "epoch": 1.31,
      "learning_rate": 7.554274993299384e-06,
      "loss": 6.1786,
      "step": 119500
    },
    {
      "epoch": 1.31,
      "learning_rate": 7.54310729920486e-06,
      "loss": 6.1809,
      "step": 120000
    },
    {
      "epoch": 1.31,
      "eval_accuracy": 0.14407839415948037,
      "eval_loss": 6.182915687561035,
      "eval_runtime": 766.7552,
      "eval_samples_per_second": 402.157,
      "eval_steps_per_second": 6.285,
      "step": 120000
    },
    {
      "epoch": 1.32,
      "learning_rate": 7.531939605110338e-06,
      "loss": 6.1831,
      "step": 120500
    },
    {
      "epoch": 1.32,
      "learning_rate": 7.520771911015814e-06,
      "loss": 6.1834,
      "step": 121000
    },
    {
      "epoch": 1.33,
      "learning_rate": 7.50960421692129e-06,
      "loss": 6.1839,
      "step": 121500
    },
    {
      "epoch": 1.33,
      "learning_rate": 7.498436522826768e-06,
      "loss": 6.1833,
      "step": 122000
    },
    {
      "epoch": 1.34,
      "learning_rate": 7.487268828732243e-06,
      "loss": 6.1767,
      "step": 122500
    },
    {
      "epoch": 1.34,
      "learning_rate": 7.476101134637721e-06,
      "loss": 6.1784,
      "step": 123000
    },
    {
      "epoch": 1.35,
      "learning_rate": 7.464933440543198e-06,
      "loss": 6.1753,
      "step": 123500
    },
    {
      "epoch": 1.35,
      "learning_rate": 7.453765746448673e-06,
      "loss": 6.178,
      "step": 124000
    },
    {
      "epoch": 1.36,
      "learning_rate": 7.442598052354151e-06,
      "loss": 6.1792,
      "step": 124500
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.431430358259627e-06,
      "loss": 6.176,
      "step": 125000
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.4202626641651045e-06,
      "loss": 6.1824,
      "step": 125500
    },
    {
      "epoch": 1.38,
      "learning_rate": 7.409094970070581e-06,
      "loss": 6.1764,
      "step": 126000
    },
    {
      "epoch": 1.38,
      "learning_rate": 7.397927275976057e-06,
      "loss": 6.1797,
      "step": 126500
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.386759581881534e-06,
      "loss": 6.1711,
      "step": 127000
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.3755918877870106e-06,
      "loss": 6.1722,
      "step": 127500
    },
    {
      "epoch": 1.4,
      "learning_rate": 7.364424193692487e-06,
      "loss": 6.1737,
      "step": 128000
    },
    {
      "epoch": 1.4,
      "learning_rate": 7.353256499597964e-06,
      "loss": 6.1692,
      "step": 128500
    },
    {
      "epoch": 1.41,
      "learning_rate": 7.34208880550344e-06,
      "loss": 6.1678,
      "step": 129000
    },
    {
      "epoch": 1.41,
      "learning_rate": 7.3309211114089174e-06,
      "loss": 6.1733,
      "step": 129500
    },
    {
      "epoch": 1.42,
      "learning_rate": 7.319753417314394e-06,
      "loss": 6.1699,
      "step": 130000
    },
    {
      "epoch": 1.42,
      "eval_accuracy": 0.1454825473557042,
      "eval_loss": 6.169208526611328,
      "eval_runtime": 765.7802,
      "eval_samples_per_second": 402.669,
      "eval_steps_per_second": 6.293,
      "step": 130000
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.30858572321987e-06,
      "loss": 6.1704,
      "step": 130500
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.297418029125347e-06,
      "loss": 6.1643,
      "step": 131000
    },
    {
      "epoch": 1.44,
      "learning_rate": 7.2862503350308235e-06,
      "loss": 6.1674,
      "step": 131500
    },
    {
      "epoch": 1.44,
      "learning_rate": 7.2750826409363e-06,
      "loss": 6.1677,
      "step": 132000
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.263914946841777e-06,
      "loss": 6.1637,
      "step": 132500
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.252747252747253e-06,
      "loss": 6.1632,
      "step": 133000
    },
    {
      "epoch": 1.46,
      "learning_rate": 7.24157955865273e-06,
      "loss": 6.164,
      "step": 133500
    },
    {
      "epoch": 1.46,
      "learning_rate": 7.230411864558207e-06,
      "loss": 6.1661,
      "step": 134000
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.219244170463683e-06,
      "loss": 6.1622,
      "step": 134500
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.20807647636916e-06,
      "loss": 6.1636,
      "step": 135000
    },
    {
      "epoch": 1.48,
      "learning_rate": 7.196908782274636e-06,
      "loss": 6.1584,
      "step": 135500
    },
    {
      "epoch": 1.49,
      "learning_rate": 7.185741088180113e-06,
      "loss": 6.1553,
      "step": 136000
    },
    {
      "epoch": 1.49,
      "learning_rate": 7.17457339408559e-06,
      "loss": 6.1627,
      "step": 136500
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.163405699991066e-06,
      "loss": 6.1637,
      "step": 137000
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.152238005896543e-06,
      "loss": 6.1597,
      "step": 137500
    },
    {
      "epoch": 1.51,
      "learning_rate": 7.14107031180202e-06,
      "loss": 6.1645,
      "step": 138000
    },
    {
      "epoch": 1.51,
      "learning_rate": 7.129902617707496e-06,
      "loss": 6.1585,
      "step": 138500
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.118734923612973e-06,
      "loss": 6.1613,
      "step": 139000
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.107567229518449e-06,
      "loss": 6.1577,
      "step": 139500
    },
    {
      "epoch": 1.53,
      "learning_rate": 7.0963995354239265e-06,
      "loss": 6.1623,
      "step": 140000
    },
    {
      "epoch": 1.53,
      "eval_accuracy": 0.1453400098451827,
      "eval_loss": 6.156247138977051,
      "eval_runtime": 765.3991,
      "eval_samples_per_second": 402.87,
      "eval_steps_per_second": 6.296,
      "step": 140000
    },
    {
      "epoch": 1.53,
      "learning_rate": 7.085231841329403e-06,
      "loss": 6.1545,
      "step": 140500
    },
    {
      "epoch": 1.54,
      "learning_rate": 7.074064147234879e-06,
      "loss": 6.1596,
      "step": 141000
    },
    {
      "epoch": 1.55,
      "learning_rate": 7.062896453140356e-06,
      "loss": 6.1577,
      "step": 141500
    },
    {
      "epoch": 1.55,
      "learning_rate": 7.0517287590458325e-06,
      "loss": 6.1556,
      "step": 142000
    },
    {
      "epoch": 1.56,
      "learning_rate": 7.040561064951309e-06,
      "loss": 6.1542,
      "step": 142500
    },
    {
      "epoch": 1.56,
      "learning_rate": 7.029393370856786e-06,
      "loss": 6.1578,
      "step": 143000
    },
    {
      "epoch": 1.57,
      "learning_rate": 7.018225676762262e-06,
      "loss": 6.1602,
      "step": 143500
    },
    {
      "epoch": 1.57,
      "learning_rate": 7.00705798266774e-06,
      "loss": 6.1578,
      "step": 144000
    },
    {
      "epoch": 1.58,
      "learning_rate": 6.995890288573216e-06,
      "loss": 6.1552,
      "step": 144500
    },
    {
      "epoch": 1.58,
      "learning_rate": 6.984722594478692e-06,
      "loss": 6.1501,
      "step": 145000
    },
    {
      "epoch": 1.59,
      "learning_rate": 6.97355490038417e-06,
      "loss": 6.1506,
      "step": 145500
    },
    {
      "epoch": 1.59,
      "learning_rate": 6.9623872062896455e-06,
      "loss": 6.1518,
      "step": 146000
    },
    {
      "epoch": 1.6,
      "learning_rate": 6.951219512195122e-06,
      "loss": 6.1533,
      "step": 146500
    },
    {
      "epoch": 1.61,
      "learning_rate": 6.9400518181006e-06,
      "loss": 6.1545,
      "step": 147000
    },
    {
      "epoch": 1.61,
      "learning_rate": 6.928884124006075e-06,
      "loss": 6.1475,
      "step": 147500
    },
    {
      "epoch": 1.62,
      "learning_rate": 6.917716429911553e-06,
      "loss": 6.1469,
      "step": 148000
    },
    {
      "epoch": 1.62,
      "learning_rate": 6.9065487358170295e-06,
      "loss": 6.144,
      "step": 148500
    },
    {
      "epoch": 1.63,
      "learning_rate": 6.895381041722505e-06,
      "loss": 6.1482,
      "step": 149000
    },
    {
      "epoch": 1.63,
      "learning_rate": 6.884213347627983e-06,
      "loss": 6.15,
      "step": 149500
    },
    {
      "epoch": 1.64,
      "learning_rate": 6.873045653533459e-06,
      "loss": 6.1422,
      "step": 150000
    },
    {
      "epoch": 1.64,
      "eval_accuracy": 0.14681557625340516,
      "eval_loss": 6.148014068603516,
      "eval_runtime": 766.4891,
      "eval_samples_per_second": 402.297,
      "eval_steps_per_second": 6.287,
      "step": 150000
    },
    {
      "epoch": 1.64,
      "learning_rate": 6.861877959438935e-06,
      "loss": 6.1491,
      "step": 150500
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.850710265344413e-06,
      "loss": 6.1442,
      "step": 151000
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.839542571249889e-06,
      "loss": 6.144,
      "step": 151500
    },
    {
      "epoch": 1.66,
      "learning_rate": 6.828374877155366e-06,
      "loss": 6.1509,
      "step": 152000
    },
    {
      "epoch": 1.67,
      "learning_rate": 6.817207183060842e-06,
      "loss": 6.1417,
      "step": 152500
    },
    {
      "epoch": 1.67,
      "learning_rate": 6.806039488966319e-06,
      "loss": 6.1473,
      "step": 153000
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.794871794871796e-06,
      "loss": 6.1496,
      "step": 153500
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.783704100777272e-06,
      "loss": 6.1466,
      "step": 154000
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.772536406682749e-06,
      "loss": 6.1462,
      "step": 154500
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.761368712588226e-06,
      "loss": 6.1388,
      "step": 155000
    },
    {
      "epoch": 1.7,
      "learning_rate": 6.750201018493702e-06,
      "loss": 6.1408,
      "step": 155500
    },
    {
      "epoch": 1.7,
      "learning_rate": 6.739033324399179e-06,
      "loss": 6.1422,
      "step": 156000
    },
    {
      "epoch": 1.71,
      "learning_rate": 6.727865630304655e-06,
      "loss": 6.1375,
      "step": 156500
    },
    {
      "epoch": 1.72,
      "learning_rate": 6.716697936210132e-06,
      "loss": 6.1404,
      "step": 157000
    },
    {
      "epoch": 1.72,
      "learning_rate": 6.705530242115609e-06,
      "loss": 6.1433,
      "step": 157500
    },
    {
      "epoch": 1.73,
      "learning_rate": 6.694362548021085e-06,
      "loss": 6.141,
      "step": 158000
    },
    {
      "epoch": 1.73,
      "learning_rate": 6.683194853926562e-06,
      "loss": 6.1426,
      "step": 158500
    },
    {
      "epoch": 1.74,
      "learning_rate": 6.6720271598320385e-06,
      "loss": 6.1448,
      "step": 159000
    },
    {
      "epoch": 1.74,
      "learning_rate": 6.660859465737515e-06,
      "loss": 6.1417,
      "step": 159500
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.649691771642992e-06,
      "loss": 6.1397,
      "step": 160000
    },
    {
      "epoch": 1.75,
      "eval_accuracy": 0.14684142929060237,
      "eval_loss": 6.136743068695068,
      "eval_runtime": 765.8201,
      "eval_samples_per_second": 402.648,
      "eval_steps_per_second": 6.293,
      "step": 160000
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.638524077548468e-06,
      "loss": 6.1416,
      "step": 160500
    },
    {
      "epoch": 1.76,
      "learning_rate": 6.6273563834539446e-06,
      "loss": 6.133,
      "step": 161000
    },
    {
      "epoch": 1.76,
      "learning_rate": 6.616188689359422e-06,
      "loss": 6.1394,
      "step": 161500
    },
    {
      "epoch": 1.77,
      "learning_rate": 6.605020995264898e-06,
      "loss": 6.1343,
      "step": 162000
    },
    {
      "epoch": 1.78,
      "learning_rate": 6.593853301170375e-06,
      "loss": 6.1429,
      "step": 162500
    },
    {
      "epoch": 1.78,
      "learning_rate": 6.5826856070758515e-06,
      "loss": 6.135,
      "step": 163000
    },
    {
      "epoch": 1.79,
      "learning_rate": 6.571517912981328e-06,
      "loss": 6.1364,
      "step": 163500
    },
    {
      "epoch": 1.79,
      "learning_rate": 6.560350218886805e-06,
      "loss": 6.1312,
      "step": 164000
    },
    {
      "epoch": 1.8,
      "learning_rate": 6.549182524792281e-06,
      "loss": 6.1332,
      "step": 164500
    },
    {
      "epoch": 1.8,
      "learning_rate": 6.5380148306977575e-06,
      "loss": 6.1319,
      "step": 165000
    },
    {
      "epoch": 1.81,
      "learning_rate": 6.526847136603235e-06,
      "loss": 6.1347,
      "step": 165500
    },
    {
      "epoch": 1.81,
      "learning_rate": 6.515679442508711e-06,
      "loss": 6.1354,
      "step": 166000
    },
    {
      "epoch": 1.82,
      "learning_rate": 6.504511748414188e-06,
      "loss": 6.1311,
      "step": 166500
    },
    {
      "epoch": 1.82,
      "learning_rate": 6.493344054319664e-06,
      "loss": 6.1336,
      "step": 167000
    },
    {
      "epoch": 1.83,
      "learning_rate": 6.482176360225141e-06,
      "loss": 6.1309,
      "step": 167500
    },
    {
      "epoch": 1.84,
      "learning_rate": 6.471008666130618e-06,
      "loss": 6.1313,
      "step": 168000
    },
    {
      "epoch": 1.84,
      "learning_rate": 6.459840972036094e-06,
      "loss": 6.1291,
      "step": 168500
    },
    {
      "epoch": 1.85,
      "learning_rate": 6.4486732779415704e-06,
      "loss": 6.1255,
      "step": 169000
    },
    {
      "epoch": 1.85,
      "learning_rate": 6.4375055838470476e-06,
      "loss": 6.1318,
      "step": 169500
    },
    {
      "epoch": 1.86,
      "learning_rate": 6.426337889752524e-06,
      "loss": 6.1342,
      "step": 170000
    },
    {
      "epoch": 1.86,
      "eval_accuracy": 0.14750071781844382,
      "eval_loss": 6.128385543823242,
      "eval_runtime": 768.3451,
      "eval_samples_per_second": 401.325,
      "eval_steps_per_second": 6.272,
      "step": 170000
    },
    {
      "epoch": 1.86,
      "learning_rate": 6.415170195658002e-06,
      "loss": 6.1251,
      "step": 170500
    },
    {
      "epoch": 1.87,
      "learning_rate": 6.404002501563477e-06,
      "loss": 6.1278,
      "step": 171000
    },
    {
      "epoch": 1.87,
      "learning_rate": 6.392834807468954e-06,
      "loss": 6.1244,
      "step": 171500
    },
    {
      "epoch": 1.88,
      "learning_rate": 6.381667113374432e-06,
      "loss": 6.1318,
      "step": 172000
    },
    {
      "epoch": 1.88,
      "learning_rate": 6.370499419279907e-06,
      "loss": 6.1314,
      "step": 172500
    },
    {
      "epoch": 1.89,
      "learning_rate": 6.359331725185385e-06,
      "loss": 6.1209,
      "step": 173000
    },
    {
      "epoch": 1.9,
      "learning_rate": 6.348164031090861e-06,
      "loss": 6.1204,
      "step": 173500
    },
    {
      "epoch": 1.9,
      "learning_rate": 6.336996336996337e-06,
      "loss": 6.1253,
      "step": 174000
    },
    {
      "epoch": 1.91,
      "learning_rate": 6.325828642901815e-06,
      "loss": 6.1224,
      "step": 174500
    },
    {
      "epoch": 1.91,
      "learning_rate": 6.314660948807291e-06,
      "loss": 6.1286,
      "step": 175000
    },
    {
      "epoch": 1.92,
      "learning_rate": 6.3034932547127665e-06,
      "loss": 6.1284,
      "step": 175500
    },
    {
      "epoch": 1.92,
      "learning_rate": 6.2923255606182445e-06,
      "loss": 6.1353,
      "step": 176000
    },
    {
      "epoch": 1.93,
      "learning_rate": 6.281157866523721e-06,
      "loss": 6.1264,
      "step": 176500
    },
    {
      "epoch": 1.93,
      "learning_rate": 6.269990172429198e-06,
      "loss": 6.1212,
      "step": 177000
    },
    {
      "epoch": 1.94,
      "learning_rate": 6.258822478334674e-06,
      "loss": 6.1183,
      "step": 177500
    },
    {
      "epoch": 1.94,
      "learning_rate": 6.2476547842401506e-06,
      "loss": 6.1241,
      "step": 178000
    },
    {
      "epoch": 1.95,
      "learning_rate": 6.236487090145628e-06,
      "loss": 6.1168,
      "step": 178500
    },
    {
      "epoch": 1.96,
      "learning_rate": 6.225319396051104e-06,
      "loss": 6.1235,
      "step": 179000
    },
    {
      "epoch": 1.96,
      "learning_rate": 6.21415170195658e-06,
      "loss": 6.125,
      "step": 179500
    },
    {
      "epoch": 1.97,
      "learning_rate": 6.2029840078620575e-06,
      "loss": 6.1291,
      "step": 180000
    },
    {
      "epoch": 1.97,
      "eval_accuracy": 0.14779425925553805,
      "eval_loss": 6.121397495269775,
      "eval_runtime": 766.405,
      "eval_samples_per_second": 402.341,
      "eval_steps_per_second": 6.288,
      "step": 180000
    },
    {
      "epoch": 1.97,
      "learning_rate": 6.191816313767534e-06,
      "loss": 6.1218,
      "step": 180500
    },
    {
      "epoch": 1.98,
      "learning_rate": 6.180648619673011e-06,
      "loss": 6.1257,
      "step": 181000
    },
    {
      "epoch": 1.98,
      "learning_rate": 6.169480925578487e-06,
      "loss": 6.1184,
      "step": 181500
    },
    {
      "epoch": 1.99,
      "learning_rate": 6.1583132314839635e-06,
      "loss": 6.118,
      "step": 182000
    },
    {
      "epoch": 1.99,
      "learning_rate": 6.147145537389441e-06,
      "loss": 6.124,
      "step": 182500
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.135977843294917e-06,
      "loss": 6.1228,
      "step": 183000
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.124810149200393e-06,
      "loss": 6.1213,
      "step": 183500
    },
    {
      "epoch": 2.01,
      "learning_rate": 6.11364245510587e-06,
      "loss": 6.1214,
      "step": 184000
    },
    {
      "epoch": 2.02,
      "learning_rate": 6.102474761011347e-06,
      "loss": 6.1215,
      "step": 184500
    },
    {
      "epoch": 2.02,
      "learning_rate": 6.091307066916824e-06,
      "loss": 6.1178,
      "step": 185000
    },
    {
      "epoch": 2.03,
      "learning_rate": 6.0801393728223e-06,
      "loss": 6.1194,
      "step": 185500
    },
    {
      "epoch": 2.03,
      "learning_rate": 6.0689716787277764e-06,
      "loss": 6.1135,
      "step": 186000
    },
    {
      "epoch": 2.04,
      "learning_rate": 6.057803984633254e-06,
      "loss": 6.1161,
      "step": 186500
    },
    {
      "epoch": 2.04,
      "learning_rate": 6.04663629053873e-06,
      "loss": 6.1232,
      "step": 187000
    },
    {
      "epoch": 2.05,
      "learning_rate": 6.035468596444207e-06,
      "loss": 6.1146,
      "step": 187500
    },
    {
      "epoch": 2.05,
      "learning_rate": 6.024300902349683e-06,
      "loss": 6.1114,
      "step": 188000
    },
    {
      "epoch": 2.06,
      "learning_rate": 6.01313320825516e-06,
      "loss": 6.1156,
      "step": 188500
    },
    {
      "epoch": 2.06,
      "learning_rate": 6.001965514160637e-06,
      "loss": 6.1192,
      "step": 189000
    },
    {
      "epoch": 2.07,
      "learning_rate": 5.990797820066113e-06,
      "loss": 6.1146,
      "step": 189500
    },
    {
      "epoch": 2.08,
      "learning_rate": 5.979630125971589e-06,
      "loss": 6.1157,
      "step": 190000
    },
    {
      "epoch": 2.08,
      "eval_accuracy": 0.14830882001884677,
      "eval_loss": 6.1131696701049805,
      "eval_runtime": 768.1193,
      "eval_samples_per_second": 401.443,
      "eval_steps_per_second": 6.274,
      "step": 190000
    },
    {
      "epoch": 2.08,
      "learning_rate": 5.9684624318770665e-06,
      "loss": 6.1136,
      "step": 190500
    },
    {
      "epoch": 2.09,
      "learning_rate": 5.957294737782543e-06,
      "loss": 6.1119,
      "step": 191000
    },
    {
      "epoch": 2.09,
      "learning_rate": 5.94612704368802e-06,
      "loss": 6.1148,
      "step": 191500
    },
    {
      "epoch": 2.1,
      "learning_rate": 5.934959349593496e-06,
      "loss": 6.1161,
      "step": 192000
    },
    {
      "epoch": 2.1,
      "learning_rate": 5.9237916554989725e-06,
      "loss": 6.1163,
      "step": 192500
    },
    {
      "epoch": 2.11,
      "learning_rate": 5.91262396140445e-06,
      "loss": 6.1122,
      "step": 193000
    },
    {
      "epoch": 2.11,
      "learning_rate": 5.901456267309926e-06,
      "loss": 6.1212,
      "step": 193500
    },
    {
      "epoch": 2.12,
      "learning_rate": 5.890288573215402e-06,
      "loss": 6.1154,
      "step": 194000
    },
    {
      "epoch": 2.12,
      "learning_rate": 5.8791208791208794e-06,
      "loss": 6.1152,
      "step": 194500
    },
    {
      "epoch": 2.13,
      "learning_rate": 5.867953185026356e-06,
      "loss": 6.1154,
      "step": 195000
    },
    {
      "epoch": 2.14,
      "learning_rate": 5.856785490931834e-06,
      "loss": 6.108,
      "step": 195500
    },
    {
      "epoch": 2.14,
      "learning_rate": 5.845617796837309e-06,
      "loss": 6.1136,
      "step": 196000
    },
    {
      "epoch": 2.15,
      "learning_rate": 5.8344501027427855e-06,
      "loss": 6.1119,
      "step": 196500
    },
    {
      "epoch": 2.15,
      "learning_rate": 5.8232824086482635e-06,
      "loss": 6.1095,
      "step": 197000
    },
    {
      "epoch": 2.16,
      "learning_rate": 5.812114714553739e-06,
      "loss": 6.1099,
      "step": 197500
    },
    {
      "epoch": 2.16,
      "learning_rate": 5.800947020459215e-06,
      "loss": 6.1151,
      "step": 198000
    },
    {
      "epoch": 2.17,
      "learning_rate": 5.789779326364693e-06,
      "loss": 6.1038,
      "step": 198500
    },
    {
      "epoch": 2.17,
      "learning_rate": 5.778611632270169e-06,
      "loss": 6.1081,
      "step": 199000
    },
    {
      "epoch": 2.18,
      "learning_rate": 5.767443938175647e-06,
      "loss": 6.1111,
      "step": 199500
    },
    {
      "epoch": 2.18,
      "learning_rate": 5.756276244081123e-06,
      "loss": 6.1146,
      "step": 200000
    },
    {
      "epoch": 2.18,
      "eval_accuracy": 0.14840016072339932,
      "eval_loss": 6.109365940093994,
      "eval_runtime": 766.2045,
      "eval_samples_per_second": 402.446,
      "eval_steps_per_second": 6.289,
      "step": 200000
    },
    {
      "epoch": 2.19,
      "learning_rate": 5.745108549986599e-06,
      "loss": 6.1092,
      "step": 200500
    },
    {
      "epoch": 2.2,
      "learning_rate": 5.733940855892076e-06,
      "loss": 6.1069,
      "step": 201000
    },
    {
      "epoch": 2.2,
      "learning_rate": 5.722773161797553e-06,
      "loss": 6.102,
      "step": 201500
    },
    {
      "epoch": 2.21,
      "learning_rate": 5.711605467703029e-06,
      "loss": 6.1034,
      "step": 202000
    },
    {
      "epoch": 2.21,
      "learning_rate": 5.700437773608506e-06,
      "loss": 6.1097,
      "step": 202500
    },
    {
      "epoch": 2.22,
      "learning_rate": 5.6892700795139824e-06,
      "loss": 6.1078,
      "step": 203000
    },
    {
      "epoch": 2.22,
      "learning_rate": 5.67810238541946e-06,
      "loss": 6.1101,
      "step": 203500
    },
    {
      "epoch": 2.23,
      "learning_rate": 5.666934691324936e-06,
      "loss": 6.1083,
      "step": 204000
    },
    {
      "epoch": 2.23,
      "learning_rate": 5.655766997230412e-06,
      "loss": 6.1017,
      "step": 204500
    },
    {
      "epoch": 2.24,
      "learning_rate": 5.644599303135889e-06,
      "loss": 6.1117,
      "step": 205000
    },
    {
      "epoch": 2.24,
      "learning_rate": 5.633431609041366e-06,
      "loss": 6.1053,
      "step": 205500
    },
    {
      "epoch": 2.25,
      "learning_rate": 5.622263914946843e-06,
      "loss": 6.1084,
      "step": 206000
    },
    {
      "epoch": 2.26,
      "learning_rate": 5.611096220852319e-06,
      "loss": 6.11,
      "step": 206500
    },
    {
      "epoch": 2.26,
      "learning_rate": 5.599928526757795e-06,
      "loss": 6.1035,
      "step": 207000
    },
    {
      "epoch": 2.27,
      "learning_rate": 5.5887608326632725e-06,
      "loss": 6.105,
      "step": 207500
    },
    {
      "epoch": 2.27,
      "learning_rate": 5.577593138568749e-06,
      "loss": 6.106,
      "step": 208000
    },
    {
      "epoch": 2.28,
      "learning_rate": 5.566425444474225e-06,
      "loss": 6.1018,
      "step": 208500
    },
    {
      "epoch": 2.28,
      "learning_rate": 5.555257750379702e-06,
      "loss": 6.1024,
      "step": 209000
    },
    {
      "epoch": 2.29,
      "learning_rate": 5.5440900562851786e-06,
      "loss": 6.103,
      "step": 209500
    },
    {
      "epoch": 2.29,
      "learning_rate": 5.532922362190656e-06,
      "loss": 6.1018,
      "step": 210000
    },
    {
      "epoch": 2.29,
      "eval_accuracy": 0.14881106200834532,
      "eval_loss": 6.10128116607666,
      "eval_runtime": 766.0908,
      "eval_samples_per_second": 402.506,
      "eval_steps_per_second": 6.29,
      "step": 210000
    },
    {
      "epoch": 2.3,
      "learning_rate": 5.521754668096132e-06,
      "loss": 6.1025,
      "step": 210500
    },
    {
      "epoch": 2.3,
      "learning_rate": 5.510586974001608e-06,
      "loss": 6.1036,
      "step": 211000
    },
    {
      "epoch": 2.31,
      "learning_rate": 5.4994192799070854e-06,
      "loss": 6.099,
      "step": 211500
    },
    {
      "epoch": 2.32,
      "learning_rate": 5.488251585812562e-06,
      "loss": 6.1007,
      "step": 212000
    },
    {
      "epoch": 2.32,
      "learning_rate": 5.477083891718038e-06,
      "loss": 6.1061,
      "step": 212500
    },
    {
      "epoch": 2.33,
      "learning_rate": 5.465916197623515e-06,
      "loss": 6.1023,
      "step": 213000
    },
    {
      "epoch": 2.33,
      "learning_rate": 5.4547485035289915e-06,
      "loss": 6.103,
      "step": 213500
    },
    {
      "epoch": 2.34,
      "learning_rate": 5.443580809434469e-06,
      "loss": 6.0991,
      "step": 214000
    },
    {
      "epoch": 2.34,
      "learning_rate": 5.432413115339945e-06,
      "loss": 6.0985,
      "step": 214500
    },
    {
      "epoch": 2.35,
      "learning_rate": 5.421245421245421e-06,
      "loss": 6.1006,
      "step": 215000
    },
    {
      "epoch": 2.35,
      "learning_rate": 5.410077727150898e-06,
      "loss": 6.0977,
      "step": 215500
    },
    {
      "epoch": 2.36,
      "learning_rate": 5.398910033056375e-06,
      "loss": 6.0988,
      "step": 216000
    },
    {
      "epoch": 2.36,
      "learning_rate": 5.387742338961851e-06,
      "loss": 6.0978,
      "step": 216500
    },
    {
      "epoch": 2.37,
      "learning_rate": 5.376574644867328e-06,
      "loss": 6.0981,
      "step": 217000
    },
    {
      "epoch": 2.38,
      "learning_rate": 5.365406950772804e-06,
      "loss": 6.0982,
      "step": 217500
    },
    {
      "epoch": 2.38,
      "learning_rate": 5.3542392566782816e-06,
      "loss": 6.1005,
      "step": 218000
    },
    {
      "epoch": 2.39,
      "learning_rate": 5.343071562583758e-06,
      "loss": 6.0987,
      "step": 218500
    },
    {
      "epoch": 2.39,
      "learning_rate": 5.331903868489234e-06,
      "loss": 6.0964,
      "step": 219000
    },
    {
      "epoch": 2.4,
      "learning_rate": 5.320736174394711e-06,
      "loss": 6.0971,
      "step": 219500
    },
    {
      "epoch": 2.4,
      "learning_rate": 5.309568480300188e-06,
      "loss": 6.1014,
      "step": 220000
    },
    {
      "epoch": 2.4,
      "eval_accuracy": 0.14876977014205314,
      "eval_loss": 6.0979390144348145,
      "eval_runtime": 766.5454,
      "eval_samples_per_second": 402.267,
      "eval_steps_per_second": 6.287,
      "step": 220000
    },
    {
      "epoch": 2.41,
      "learning_rate": 5.298400786205666e-06,
      "loss": 6.0934,
      "step": 220500
    },
    {
      "epoch": 2.41,
      "learning_rate": 5.287233092111141e-06,
      "loss": 6.0987,
      "step": 221000
    },
    {
      "epoch": 2.42,
      "learning_rate": 5.276065398016617e-06,
      "loss": 6.1008,
      "step": 221500
    },
    {
      "epoch": 2.43,
      "learning_rate": 5.264897703922095e-06,
      "loss": 6.0847,
      "step": 222000
    },
    {
      "epoch": 2.43,
      "learning_rate": 5.253730009827571e-06,
      "loss": 6.0959,
      "step": 222500
    },
    {
      "epoch": 2.44,
      "learning_rate": 5.242562315733047e-06,
      "loss": 6.0913,
      "step": 223000
    },
    {
      "epoch": 2.44,
      "learning_rate": 5.231394621638525e-06,
      "loss": 6.0922,
      "step": 223500
    },
    {
      "epoch": 2.45,
      "learning_rate": 5.220226927544001e-06,
      "loss": 6.0995,
      "step": 224000
    },
    {
      "epoch": 2.45,
      "learning_rate": 5.2090592334494785e-06,
      "loss": 6.1,
      "step": 224500
    },
    {
      "epoch": 2.46,
      "learning_rate": 5.197891539354955e-06,
      "loss": 6.0999,
      "step": 225000
    },
    {
      "epoch": 2.46,
      "learning_rate": 5.186723845260431e-06,
      "loss": 6.0957,
      "step": 225500
    },
    {
      "epoch": 2.47,
      "learning_rate": 5.175556151165908e-06,
      "loss": 6.0934,
      "step": 226000
    },
    {
      "epoch": 2.47,
      "learning_rate": 5.1643884570713846e-06,
      "loss": 6.0998,
      "step": 226500
    },
    {
      "epoch": 2.48,
      "learning_rate": 5.153220762976861e-06,
      "loss": 6.0959,
      "step": 227000
    },
    {
      "epoch": 2.49,
      "learning_rate": 5.142053068882338e-06,
      "loss": 6.0955,
      "step": 227500
    },
    {
      "epoch": 2.49,
      "learning_rate": 5.130885374787814e-06,
      "loss": 6.0905,
      "step": 228000
    },
    {
      "epoch": 2.5,
      "learning_rate": 5.1197176806932914e-06,
      "loss": 6.1,
      "step": 228500
    },
    {
      "epoch": 2.5,
      "learning_rate": 5.108549986598768e-06,
      "loss": 6.0907,
      "step": 229000
    },
    {
      "epoch": 2.51,
      "learning_rate": 5.097382292504244e-06,
      "loss": 6.0914,
      "step": 229500
    },
    {
      "epoch": 2.51,
      "learning_rate": 5.086214598409721e-06,
      "loss": 6.0935,
      "step": 230000
    },
    {
      "epoch": 2.51,
      "eval_accuracy": 0.1489331512830438,
      "eval_loss": 6.093608379364014,
      "eval_runtime": 765.9143,
      "eval_samples_per_second": 402.599,
      "eval_steps_per_second": 6.292,
      "step": 230000
    },
    {
      "epoch": 2.52,
      "learning_rate": 5.0750469043151975e-06,
      "loss": 6.0975,
      "step": 230500
    },
    {
      "epoch": 2.52,
      "learning_rate": 5.063879210220674e-06,
      "loss": 6.0952,
      "step": 231000
    },
    {
      "epoch": 2.53,
      "learning_rate": 5.052711516126151e-06,
      "loss": 6.0944,
      "step": 231500
    },
    {
      "epoch": 2.53,
      "learning_rate": 5.041543822031627e-06,
      "loss": 6.0949,
      "step": 232000
    },
    {
      "epoch": 2.54,
      "learning_rate": 5.030376127937104e-06,
      "loss": 6.0938,
      "step": 232500
    },
    {
      "epoch": 2.55,
      "learning_rate": 5.019208433842581e-06,
      "loss": 6.0978,
      "step": 233000
    },
    {
      "epoch": 2.55,
      "learning_rate": 5.008040739748057e-06,
      "loss": 6.0889,
      "step": 233500
    },
    {
      "epoch": 2.56,
      "learning_rate": 4.996873045653533e-06,
      "loss": 6.0874,
      "step": 234000
    },
    {
      "epoch": 2.56,
      "learning_rate": 4.98570535155901e-06,
      "loss": 6.0947,
      "step": 234500
    },
    {
      "epoch": 2.57,
      "learning_rate": 4.9745376574644876e-06,
      "loss": 6.0922,
      "step": 235000
    },
    {
      "epoch": 2.57,
      "learning_rate": 4.963369963369964e-06,
      "loss": 6.0901,
      "step": 235500
    },
    {
      "epoch": 2.58,
      "learning_rate": 4.952202269275441e-06,
      "loss": 6.0893,
      "step": 236000
    },
    {
      "epoch": 2.58,
      "learning_rate": 4.941034575180917e-06,
      "loss": 6.093,
      "step": 236500
    },
    {
      "epoch": 2.59,
      "learning_rate": 4.929866881086394e-06,
      "loss": 6.0937,
      "step": 237000
    },
    {
      "epoch": 2.59,
      "learning_rate": 4.918699186991871e-06,
      "loss": 6.0911,
      "step": 237500
    },
    {
      "epoch": 2.6,
      "learning_rate": 4.907531492897347e-06,
      "loss": 6.0927,
      "step": 238000
    },
    {
      "epoch": 2.61,
      "learning_rate": 4.896363798802823e-06,
      "loss": 6.0887,
      "step": 238500
    },
    {
      "epoch": 2.61,
      "learning_rate": 4.8851961047083005e-06,
      "loss": 6.0924,
      "step": 239000
    },
    {
      "epoch": 2.62,
      "learning_rate": 4.874028410613777e-06,
      "loss": 6.0873,
      "step": 239500
    },
    {
      "epoch": 2.62,
      "learning_rate": 4.862860716519254e-06,
      "loss": 6.0899,
      "step": 240000
    },
    {
      "epoch": 2.62,
      "eval_accuracy": 0.14914646441496213,
      "eval_loss": 6.088108539581299,
      "eval_runtime": 766.5214,
      "eval_samples_per_second": 402.28,
      "eval_steps_per_second": 6.287,
      "step": 240000
    },
    {
      "epoch": 2.63,
      "learning_rate": 4.85169302242473e-06,
      "loss": 6.0858,
      "step": 240500
    },
    {
      "epoch": 2.63,
      "learning_rate": 4.8405253283302065e-06,
      "loss": 6.0905,
      "step": 241000
    },
    {
      "epoch": 2.64,
      "learning_rate": 4.829357634235684e-06,
      "loss": 6.0856,
      "step": 241500
    },
    {
      "epoch": 2.64,
      "learning_rate": 4.81818994014116e-06,
      "loss": 6.0876,
      "step": 242000
    },
    {
      "epoch": 2.65,
      "learning_rate": 4.807022246046636e-06,
      "loss": 6.0895,
      "step": 242500
    },
    {
      "epoch": 2.65,
      "learning_rate": 4.795854551952113e-06,
      "loss": 6.0876,
      "step": 243000
    },
    {
      "epoch": 2.66,
      "learning_rate": 4.78468685785759e-06,
      "loss": 6.0832,
      "step": 243500
    },
    {
      "epoch": 2.67,
      "learning_rate": 4.773519163763067e-06,
      "loss": 6.0864,
      "step": 244000
    },
    {
      "epoch": 2.67,
      "learning_rate": 4.762351469668543e-06,
      "loss": 6.0893,
      "step": 244500
    },
    {
      "epoch": 2.68,
      "learning_rate": 4.7511837755740195e-06,
      "loss": 6.0861,
      "step": 245000
    },
    {
      "epoch": 2.68,
      "learning_rate": 4.740016081479497e-06,
      "loss": 6.0878,
      "step": 245500
    },
    {
      "epoch": 2.69,
      "learning_rate": 4.728848387384973e-06,
      "loss": 6.0861,
      "step": 246000
    },
    {
      "epoch": 2.69,
      "learning_rate": 4.717680693290449e-06,
      "loss": 6.0812,
      "step": 246500
    },
    {
      "epoch": 2.7,
      "learning_rate": 4.706512999195926e-06,
      "loss": 6.0853,
      "step": 247000
    },
    {
      "epoch": 2.7,
      "learning_rate": 4.6953453051014035e-06,
      "loss": 6.0809,
      "step": 247500
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.68417761100688e-06,
      "loss": 6.0891,
      "step": 248000
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.673009916912356e-06,
      "loss": 6.0882,
      "step": 248500
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.661842222817833e-06,
      "loss": 6.0836,
      "step": 249000
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.6506745287233095e-06,
      "loss": 6.0892,
      "step": 249500
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.639506834628787e-06,
      "loss": 6.0858,
      "step": 250000
    },
    {
      "epoch": 2.73,
      "eval_accuracy": 0.14981958424885708,
      "eval_loss": 6.0851149559021,
      "eval_runtime": 766.4989,
      "eval_samples_per_second": 402.291,
      "eval_steps_per_second": 6.287,
      "step": 250000
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.628339140534263e-06,
      "loss": 6.0871,
      "step": 250500
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.617171446439739e-06,
      "loss": 6.0809,
      "step": 251000
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.606003752345216e-06,
      "loss": 6.08,
      "step": 251500
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.594836058250693e-06,
      "loss": 6.0815,
      "step": 252000
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.58366836415617e-06,
      "loss": 6.0868,
      "step": 252500
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.572500670061646e-06,
      "loss": 6.0815,
      "step": 253000
    },
    {
      "epoch": 2.77,
      "learning_rate": 4.5613329759671225e-06,
      "loss": 6.0848,
      "step": 253500
    },
    {
      "epoch": 2.77,
      "learning_rate": 4.5501652818726e-06,
      "loss": 6.0802,
      "step": 254000
    },
    {
      "epoch": 2.78,
      "learning_rate": 4.538997587778076e-06,
      "loss": 6.085,
      "step": 254500
    },
    {
      "epoch": 2.79,
      "learning_rate": 4.527829893683552e-06,
      "loss": 6.0849,
      "step": 255000
    },
    {
      "epoch": 2.79,
      "learning_rate": 4.516662199589029e-06,
      "loss": 6.0796,
      "step": 255500
    },
    {
      "epoch": 2.8,
      "learning_rate": 4.505494505494506e-06,
      "loss": 6.0838,
      "step": 256000
    },
    {
      "epoch": 2.8,
      "learning_rate": 4.494326811399983e-06,
      "loss": 6.0819,
      "step": 256500
    },
    {
      "epoch": 2.81,
      "learning_rate": 4.483159117305459e-06,
      "loss": 6.0798,
      "step": 257000
    },
    {
      "epoch": 2.81,
      "learning_rate": 4.471991423210935e-06,
      "loss": 6.0765,
      "step": 257500
    },
    {
      "epoch": 2.82,
      "learning_rate": 4.4608237291164125e-06,
      "loss": 6.0813,
      "step": 258000
    },
    {
      "epoch": 2.82,
      "learning_rate": 4.44965603502189e-06,
      "loss": 6.0841,
      "step": 258500
    },
    {
      "epoch": 2.83,
      "learning_rate": 4.438488340927365e-06,
      "loss": 6.0809,
      "step": 259000
    },
    {
      "epoch": 2.83,
      "learning_rate": 4.427320646832842e-06,
      "loss": 6.0849,
      "step": 259500
    },
    {
      "epoch": 2.84,
      "learning_rate": 4.416152952738319e-06,
      "loss": 6.0872,
      "step": 260000
    },
    {
      "epoch": 2.84,
      "eval_accuracy": 0.14966085941523033,
      "eval_loss": 6.081911563873291,
      "eval_runtime": 766.0948,
      "eval_samples_per_second": 402.504,
      "eval_steps_per_second": 6.29,
      "step": 260000
    },
    {
      "epoch": 2.85,
      "learning_rate": 4.404985258643796e-06,
      "loss": 6.0818,
      "step": 260500
    },
    {
      "epoch": 2.85,
      "learning_rate": 4.393817564549272e-06,
      "loss": 6.0804,
      "step": 261000
    },
    {
      "epoch": 2.86,
      "learning_rate": 4.382649870454749e-06,
      "loss": 6.0829,
      "step": 261500
    },
    {
      "epoch": 2.86,
      "learning_rate": 4.3714821763602255e-06,
      "loss": 6.0891,
      "step": 262000
    },
    {
      "epoch": 2.87,
      "learning_rate": 4.360314482265703e-06,
      "loss": 6.0821,
      "step": 262500
    },
    {
      "epoch": 2.87,
      "learning_rate": 4.349146788171179e-06,
      "loss": 6.0787,
      "step": 263000
    },
    {
      "epoch": 2.88,
      "learning_rate": 4.337979094076655e-06,
      "loss": 6.0761,
      "step": 263500
    },
    {
      "epoch": 2.88,
      "learning_rate": 4.326811399982132e-06,
      "loss": 6.0836,
      "step": 264000
    },
    {
      "epoch": 2.89,
      "learning_rate": 4.315643705887609e-06,
      "loss": 6.0825,
      "step": 264500
    },
    {
      "epoch": 2.89,
      "learning_rate": 4.304476011793085e-06,
      "loss": 6.0806,
      "step": 265000
    },
    {
      "epoch": 2.9,
      "learning_rate": 4.293308317698562e-06,
      "loss": 6.0727,
      "step": 265500
    },
    {
      "epoch": 2.91,
      "learning_rate": 4.282140623604038e-06,
      "loss": 6.0841,
      "step": 266000
    },
    {
      "epoch": 2.91,
      "learning_rate": 4.2709729295095155e-06,
      "loss": 6.0756,
      "step": 266500
    },
    {
      "epoch": 2.92,
      "learning_rate": 4.259805235414992e-06,
      "loss": 6.0775,
      "step": 267000
    },
    {
      "epoch": 2.92,
      "learning_rate": 4.248637541320468e-06,
      "loss": 6.0831,
      "step": 267500
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.237469847225945e-06,
      "loss": 6.078,
      "step": 268000
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.2263021531314216e-06,
      "loss": 6.0834,
      "step": 268500
    },
    {
      "epoch": 2.94,
      "learning_rate": 4.215134459036899e-06,
      "loss": 6.0759,
      "step": 269000
    },
    {
      "epoch": 2.94,
      "learning_rate": 4.203966764942375e-06,
      "loss": 6.0755,
      "step": 269500
    },
    {
      "epoch": 2.95,
      "learning_rate": 4.192799070847851e-06,
      "loss": 6.0858,
      "step": 270000
    },
    {
      "epoch": 2.95,
      "eval_accuracy": 0.1499890329360819,
      "eval_loss": 6.078388214111328,
      "eval_runtime": 767.5827,
      "eval_samples_per_second": 401.723,
      "eval_steps_per_second": 6.278,
      "step": 270000
    },
    {
      "epoch": 2.95,
      "learning_rate": 4.1816313767533285e-06,
      "loss": 6.0846,
      "step": 270500
    },
    {
      "epoch": 2.96,
      "learning_rate": 4.170463682658806e-06,
      "loss": 6.0775,
      "step": 271000
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.159295988564281e-06,
      "loss": 6.0783,
      "step": 271500
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.148128294469758e-06,
      "loss": 6.0808,
      "step": 272000
    },
    {
      "epoch": 2.98,
      "learning_rate": 4.136960600375235e-06,
      "loss": 6.0809,
      "step": 272500
    },
    {
      "epoch": 2.98,
      "learning_rate": 4.125792906280712e-06,
      "loss": 6.0724,
      "step": 273000
    },
    {
      "epoch": 2.99,
      "learning_rate": 4.114625212186188e-06,
      "loss": 6.0774,
      "step": 273500
    },
    {
      "epoch": 2.99,
      "learning_rate": 4.103457518091665e-06,
      "loss": 6.0815,
      "step": 274000
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.092289823997141e-06,
      "loss": 6.0806,
      "step": 274500
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.0811221299026185e-06,
      "loss": 6.0772,
      "step": 275000
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.069954435808095e-06,
      "loss": 6.0736,
      "step": 275500
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.058786741713571e-06,
      "loss": 6.0784,
      "step": 276000
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.047619047619048e-06,
      "loss": 6.0785,
      "step": 276500
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.0364513535245246e-06,
      "loss": 6.0697,
      "step": 277000
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.025283659430001e-06,
      "loss": 6.0752,
      "step": 277500
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.014115965335478e-06,
      "loss": 6.0769,
      "step": 278000
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.002948271240954e-06,
      "loss": 6.081,
      "step": 278500
    },
    {
      "epoch": 3.05,
      "learning_rate": 3.9917805771464315e-06,
      "loss": 6.0742,
      "step": 279000
    },
    {
      "epoch": 3.05,
      "learning_rate": 3.980612883051908e-06,
      "loss": 6.0753,
      "step": 279500
    },
    {
      "epoch": 3.06,
      "learning_rate": 3.969445188957384e-06,
      "loss": 6.0775,
      "step": 280000
    },
    {
      "epoch": 3.06,
      "eval_accuracy": 0.1500744074772058,
      "eval_loss": 6.074457168579102,
      "eval_runtime": 770.7739,
      "eval_samples_per_second": 400.06,
      "eval_steps_per_second": 6.252,
      "step": 280000
    },
    {
      "epoch": 3.06,
      "learning_rate": 3.958277494862861e-06,
      "loss": 6.0758,
      "step": 280500
    },
    {
      "epoch": 3.07,
      "learning_rate": 3.9471098007683375e-06,
      "loss": 6.0767,
      "step": 281000
    },
    {
      "epoch": 3.08,
      "learning_rate": 3.935942106673814e-06,
      "loss": 6.0749,
      "step": 281500
    },
    {
      "epoch": 3.08,
      "learning_rate": 3.924774412579291e-06,
      "loss": 6.0751,
      "step": 282000
    },
    {
      "epoch": 3.09,
      "learning_rate": 3.913606718484767e-06,
      "loss": 6.0731,
      "step": 282500
    },
    {
      "epoch": 3.09,
      "learning_rate": 3.902439024390244e-06,
      "loss": 6.077,
      "step": 283000
    },
    {
      "epoch": 3.1,
      "learning_rate": 3.8912713302957215e-06,
      "loss": 6.0744,
      "step": 283500
    },
    {
      "epoch": 3.1,
      "learning_rate": 3.880103636201197e-06,
      "loss": 6.0725,
      "step": 284000
    },
    {
      "epoch": 3.11,
      "learning_rate": 3.868935942106674e-06,
      "loss": 6.0783,
      "step": 284500
    },
    {
      "epoch": 3.11,
      "learning_rate": 3.857768248012151e-06,
      "loss": 6.0651,
      "step": 285000
    },
    {
      "epoch": 3.12,
      "learning_rate": 3.8466005539176276e-06,
      "loss": 6.0807,
      "step": 285500
    },
    {
      "epoch": 3.12,
      "learning_rate": 3.835432859823104e-06,
      "loss": 6.0759,
      "step": 286000
    },
    {
      "epoch": 3.13,
      "learning_rate": 3.824265165728581e-06,
      "loss": 6.0745,
      "step": 286500
    },
    {
      "epoch": 3.14,
      "learning_rate": 3.8130974716340573e-06,
      "loss": 6.0734,
      "step": 287000
    },
    {
      "epoch": 3.14,
      "learning_rate": 3.801929777539534e-06,
      "loss": 6.0728,
      "step": 287500
    },
    {
      "epoch": 3.15,
      "learning_rate": 3.7907620834450103e-06,
      "loss": 6.0711,
      "step": 288000
    },
    {
      "epoch": 3.15,
      "learning_rate": 3.779594389350487e-06,
      "loss": 6.0752,
      "step": 288500
    },
    {
      "epoch": 3.16,
      "learning_rate": 3.7684266952559638e-06,
      "loss": 6.0733,
      "step": 289000
    },
    {
      "epoch": 3.16,
      "learning_rate": 3.7572590011614405e-06,
      "loss": 6.074,
      "step": 289500
    },
    {
      "epoch": 3.17,
      "learning_rate": 3.746091307066917e-06,
      "loss": 6.0715,
      "step": 290000
    },
    {
      "epoch": 3.17,
      "eval_accuracy": 0.1501575925073286,
      "eval_loss": 6.072000980377197,
      "eval_runtime": 769.4522,
      "eval_samples_per_second": 400.747,
      "eval_steps_per_second": 6.263,
      "step": 290000
    },
    {
      "epoch": 3.17,
      "learning_rate": 3.7349236129723935e-06,
      "loss": 6.0716,
      "step": 290500
    },
    {
      "epoch": 3.18,
      "learning_rate": 3.7237559188778707e-06,
      "loss": 6.0735,
      "step": 291000
    },
    {
      "epoch": 3.18,
      "learning_rate": 3.7125882247833474e-06,
      "loss": 6.0726,
      "step": 291500
    },
    {
      "epoch": 3.19,
      "learning_rate": 3.7014205306888233e-06,
      "loss": 6.073,
      "step": 292000
    },
    {
      "epoch": 3.2,
      "learning_rate": 3.6902528365943004e-06,
      "loss": 6.0675,
      "step": 292500
    },
    {
      "epoch": 3.2,
      "learning_rate": 3.679085142499777e-06,
      "loss": 6.0665,
      "step": 293000
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.667917448405254e-06,
      "loss": 6.0713,
      "step": 293500
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.65674975431073e-06,
      "loss": 6.0748,
      "step": 294000
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.645582060216207e-06,
      "loss": 6.0666,
      "step": 294500
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.6344143661216836e-06,
      "loss": 6.0685,
      "step": 295000
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.6232466720271603e-06,
      "loss": 6.0705,
      "step": 295500
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.6120789779326366e-06,
      "loss": 6.0686,
      "step": 296000
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.6009112838381133e-06,
      "loss": 6.072,
      "step": 296500
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.58974358974359e-06,
      "loss": 6.0696,
      "step": 297000
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.5785758956490668e-06,
      "loss": 6.0733,
      "step": 297500
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.567408201554543e-06,
      "loss": 6.0716,
      "step": 298000
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.55624050746002e-06,
      "loss": 6.0739,
      "step": 298500
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.5450728133654965e-06,
      "loss": 6.0684,
      "step": 299000
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.5339051192709732e-06,
      "loss": 6.0665,
      "step": 299500
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.52273742517645e-06,
      "loss": 6.0704,
      "step": 300000
    },
    {
      "epoch": 3.28,
      "eval_accuracy": 0.1501824818735117,
      "eval_loss": 6.069890975952148,
      "eval_runtime": 768.119,
      "eval_samples_per_second": 401.443,
      "eval_steps_per_second": 6.274,
      "step": 300000
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.5115697310819263e-06,
      "loss": 6.0736,
      "step": 300500
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.500402036987403e-06,
      "loss": 6.0757,
      "step": 301000
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.4892343428928797e-06,
      "loss": 6.0614,
      "step": 301500
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.478066648798357e-06,
      "loss": 6.0669,
      "step": 302000
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.4668989547038327e-06,
      "loss": 6.0722,
      "step": 302500
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.4557312606093095e-06,
      "loss": 6.0732,
      "step": 303000
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.4445635665147866e-06,
      "loss": 6.074,
      "step": 303500
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.4333958724202633e-06,
      "loss": 6.0732,
      "step": 304000
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.422228178325739e-06,
      "loss": 6.07,
      "step": 304500
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.4110604842312163e-06,
      "loss": 6.0686,
      "step": 305000
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.399892790136693e-06,
      "loss": 6.0702,
      "step": 305500
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.3887250960421698e-06,
      "loss": 6.0699,
      "step": 306000
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.377557401947646e-06,
      "loss": 6.0691,
      "step": 306500
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.366389707853123e-06,
      "loss": 6.0638,
      "step": 307000
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.3552220137585995e-06,
      "loss": 6.0639,
      "step": 307500
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.3440543196640762e-06,
      "loss": 6.062,
      "step": 308000
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.3328866255695525e-06,
      "loss": 6.0624,
      "step": 308500
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.3217189314750293e-06,
      "loss": 6.0637,
      "step": 309000
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.310551237380506e-06,
      "loss": 6.0682,
      "step": 309500
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.2993835432859827e-06,
      "loss": 6.0678,
      "step": 310000
    },
    {
      "epoch": 3.39,
      "eval_accuracy": 0.15031177649197602,
      "eval_loss": 6.066825866699219,
      "eval_runtime": 766.8548,
      "eval_samples_per_second": 402.105,
      "eval_steps_per_second": 6.284,
      "step": 310000
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.288215849191459e-06,
      "loss": 6.0694,
      "step": 310500
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.2770481550969357e-06,
      "loss": 6.0619,
      "step": 311000
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.2658804610024125e-06,
      "loss": 6.0697,
      "step": 311500
    },
    {
      "epoch": 3.41,
      "learning_rate": 3.254712766907889e-06,
      "loss": 6.0766,
      "step": 312000
    },
    {
      "epoch": 3.41,
      "learning_rate": 3.2435450728133655e-06,
      "loss": 6.0705,
      "step": 312500
    },
    {
      "epoch": 3.42,
      "learning_rate": 3.232377378718842e-06,
      "loss": 6.0629,
      "step": 313000
    },
    {
      "epoch": 3.42,
      "learning_rate": 3.221209684624319e-06,
      "loss": 6.0669,
      "step": 313500
    },
    {
      "epoch": 3.43,
      "learning_rate": 3.2100419905297956e-06,
      "loss": 6.0717,
      "step": 314000
    },
    {
      "epoch": 3.44,
      "learning_rate": 3.198874296435272e-06,
      "loss": 6.067,
      "step": 314500
    },
    {
      "epoch": 3.44,
      "learning_rate": 3.1877066023407487e-06,
      "loss": 6.0658,
      "step": 315000
    },
    {
      "epoch": 3.45,
      "learning_rate": 3.1765389082462254e-06,
      "loss": 6.0667,
      "step": 315500
    },
    {
      "epoch": 3.45,
      "learning_rate": 3.1653712141517025e-06,
      "loss": 6.0698,
      "step": 316000
    },
    {
      "epoch": 3.46,
      "learning_rate": 3.1542035200571792e-06,
      "loss": 6.0616,
      "step": 316500
    },
    {
      "epoch": 3.46,
      "learning_rate": 3.143035825962655e-06,
      "loss": 6.0687,
      "step": 317000
    },
    {
      "epoch": 3.47,
      "learning_rate": 3.1318681318681323e-06,
      "loss": 6.0637,
      "step": 317500
    },
    {
      "epoch": 3.47,
      "learning_rate": 3.120700437773609e-06,
      "loss": 6.0652,
      "step": 318000
    },
    {
      "epoch": 3.48,
      "learning_rate": 3.1095327436790857e-06,
      "loss": 6.0609,
      "step": 318500
    },
    {
      "epoch": 3.48,
      "learning_rate": 3.098365049584562e-06,
      "loss": 6.0673,
      "step": 319000
    },
    {
      "epoch": 3.49,
      "learning_rate": 3.0871973554900387e-06,
      "loss": 6.0684,
      "step": 319500
    },
    {
      "epoch": 3.5,
      "learning_rate": 3.0760296613955155e-06,
      "loss": 6.0662,
      "step": 320000
    },
    {
      "epoch": 3.5,
      "eval_accuracy": 0.15030049989727468,
      "eval_loss": 6.064922332763672,
      "eval_runtime": 766.9226,
      "eval_samples_per_second": 402.069,
      "eval_steps_per_second": 6.284,
      "step": 320000
    },
    {
      "epoch": 3.5,
      "learning_rate": 3.064861967300992e-06,
      "loss": 6.0639,
      "step": 320500
    },
    {
      "epoch": 3.51,
      "learning_rate": 3.0536942732064685e-06,
      "loss": 6.0638,
      "step": 321000
    },
    {
      "epoch": 3.51,
      "learning_rate": 3.042526579111945e-06,
      "loss": 6.0633,
      "step": 321500
    },
    {
      "epoch": 3.52,
      "learning_rate": 3.031358885017422e-06,
      "loss": 6.0645,
      "step": 322000
    },
    {
      "epoch": 3.52,
      "learning_rate": 3.0201911909228986e-06,
      "loss": 6.0621,
      "step": 322500
    },
    {
      "epoch": 3.53,
      "learning_rate": 3.009023496828375e-06,
      "loss": 6.0674,
      "step": 323000
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.9978558027338517e-06,
      "loss": 6.0671,
      "step": 323500
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.9866881086393284e-06,
      "loss": 6.0607,
      "step": 324000
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.975520414544805e-06,
      "loss": 6.0693,
      "step": 324500
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.9643527204502814e-06,
      "loss": 6.0557,
      "step": 325000
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.953185026355758e-06,
      "loss": 6.0582,
      "step": 325500
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.942017332261235e-06,
      "loss": 6.0689,
      "step": 326000
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.9308496381667116e-06,
      "loss": 6.0596,
      "step": 326500
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.919681944072188e-06,
      "loss": 6.0673,
      "step": 327000
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.9085142499776646e-06,
      "loss": 6.0655,
      "step": 327500
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.8973465558831413e-06,
      "loss": 6.0642,
      "step": 328000
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.8861788617886185e-06,
      "loss": 6.0636,
      "step": 328500
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.8750111676940943e-06,
      "loss": 6.0627,
      "step": 329000
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.8638434735995715e-06,
      "loss": 6.0658,
      "step": 329500
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.852675779505048e-06,
      "loss": 6.0569,
      "step": 330000
    },
    {
      "epoch": 3.6,
      "eval_accuracy": 0.1505006922434429,
      "eval_loss": 6.062248706817627,
      "eval_runtime": 766.7908,
      "eval_samples_per_second": 402.138,
      "eval_steps_per_second": 6.285,
      "step": 330000
    },
    {
      "epoch": 3.61,
      "learning_rate": 2.841508085410525e-06,
      "loss": 6.0681,
      "step": 330500
    },
    {
      "epoch": 3.62,
      "learning_rate": 2.8303403913160016e-06,
      "loss": 6.058,
      "step": 331000
    },
    {
      "epoch": 3.62,
      "learning_rate": 2.819172697221478e-06,
      "loss": 6.0606,
      "step": 331500
    },
    {
      "epoch": 3.63,
      "learning_rate": 2.8080050031269547e-06,
      "loss": 6.063,
      "step": 332000
    },
    {
      "epoch": 3.63,
      "learning_rate": 2.7968373090324314e-06,
      "loss": 6.06,
      "step": 332500
    },
    {
      "epoch": 3.64,
      "learning_rate": 2.785669614937908e-06,
      "loss": 6.0602,
      "step": 333000
    },
    {
      "epoch": 3.64,
      "learning_rate": 2.7745019208433844e-06,
      "loss": 6.0561,
      "step": 333500
    },
    {
      "epoch": 3.65,
      "learning_rate": 2.763334226748861e-06,
      "loss": 6.0605,
      "step": 334000
    },
    {
      "epoch": 3.65,
      "learning_rate": 2.752166532654338e-06,
      "loss": 6.0672,
      "step": 334500
    },
    {
      "epoch": 3.66,
      "learning_rate": 2.7409988385598146e-06,
      "loss": 6.0624,
      "step": 335000
    },
    {
      "epoch": 3.66,
      "learning_rate": 2.729831144465291e-06,
      "loss": 6.0643,
      "step": 335500
    },
    {
      "epoch": 3.67,
      "learning_rate": 2.7186634503707676e-06,
      "loss": 6.0584,
      "step": 336000
    },
    {
      "epoch": 3.68,
      "learning_rate": 2.7074957562762443e-06,
      "loss": 6.0637,
      "step": 336500
    },
    {
      "epoch": 3.68,
      "learning_rate": 2.696328062181721e-06,
      "loss": 6.0612,
      "step": 337000
    },
    {
      "epoch": 3.69,
      "learning_rate": 2.6851603680871973e-06,
      "loss": 6.0562,
      "step": 337500
    },
    {
      "epoch": 3.69,
      "learning_rate": 2.673992673992674e-06,
      "loss": 6.0606,
      "step": 338000
    },
    {
      "epoch": 3.7,
      "learning_rate": 2.6628249798981508e-06,
      "loss": 6.0687,
      "step": 338500
    },
    {
      "epoch": 3.7,
      "learning_rate": 2.6516572858036275e-06,
      "loss": 6.0556,
      "step": 339000
    },
    {
      "epoch": 3.71,
      "learning_rate": 2.640489591709104e-06,
      "loss": 6.0653,
      "step": 339500
    },
    {
      "epoch": 3.71,
      "learning_rate": 2.6293218976145805e-06,
      "loss": 6.0604,
      "step": 340000
    },
    {
      "epoch": 3.71,
      "eval_accuracy": 0.1505791744366545,
      "eval_loss": 6.061243534088135,
      "eval_runtime": 768.1037,
      "eval_samples_per_second": 401.451,
      "eval_steps_per_second": 6.274,
      "step": 340000
    },
    {
      "epoch": 3.72,
      "learning_rate": 2.6181542035200572e-06,
      "loss": 6.0615,
      "step": 340500
    },
    {
      "epoch": 3.72,
      "learning_rate": 2.6069865094255344e-06,
      "loss": 6.0616,
      "step": 341000
    },
    {
      "epoch": 3.73,
      "learning_rate": 2.5958188153310103e-06,
      "loss": 6.0599,
      "step": 341500
    },
    {
      "epoch": 3.74,
      "learning_rate": 2.5846511212364874e-06,
      "loss": 6.0615,
      "step": 342000
    },
    {
      "epoch": 3.74,
      "learning_rate": 2.573483427141964e-06,
      "loss": 6.0647,
      "step": 342500
    },
    {
      "epoch": 3.75,
      "learning_rate": 2.562315733047441e-06,
      "loss": 6.0624,
      "step": 343000
    },
    {
      "epoch": 3.75,
      "learning_rate": 2.551148038952917e-06,
      "loss": 6.0538,
      "step": 343500
    },
    {
      "epoch": 3.76,
      "learning_rate": 2.539980344858394e-06,
      "loss": 6.0609,
      "step": 344000
    },
    {
      "epoch": 3.76,
      "learning_rate": 2.5288126507638706e-06,
      "loss": 6.0601,
      "step": 344500
    },
    {
      "epoch": 3.77,
      "learning_rate": 2.5176449566693473e-06,
      "loss": 6.0678,
      "step": 345000
    },
    {
      "epoch": 3.77,
      "learning_rate": 2.5064772625748236e-06,
      "loss": 6.0627,
      "step": 345500
    },
    {
      "epoch": 3.78,
      "learning_rate": 2.4953095684803003e-06,
      "loss": 6.0604,
      "step": 346000
    },
    {
      "epoch": 3.79,
      "learning_rate": 2.484141874385777e-06,
      "loss": 6.0543,
      "step": 346500
    },
    {
      "epoch": 3.79,
      "learning_rate": 2.4729741802912538e-06,
      "loss": 6.0597,
      "step": 347000
    },
    {
      "epoch": 3.8,
      "learning_rate": 2.4618064861967305e-06,
      "loss": 6.0545,
      "step": 347500
    },
    {
      "epoch": 3.8,
      "learning_rate": 2.4506387921022072e-06,
      "loss": 6.0584,
      "step": 348000
    },
    {
      "epoch": 3.81,
      "learning_rate": 2.4394710980076835e-06,
      "loss": 6.0626,
      "step": 348500
    },
    {
      "epoch": 3.81,
      "learning_rate": 2.4283034039131602e-06,
      "loss": 6.0635,
      "step": 349000
    },
    {
      "epoch": 3.82,
      "learning_rate": 2.417135709818637e-06,
      "loss": 6.0609,
      "step": 349500
    },
    {
      "epoch": 3.82,
      "learning_rate": 2.4059680157241137e-06,
      "loss": 6.0525,
      "step": 350000
    },
    {
      "epoch": 3.82,
      "eval_accuracy": 0.15072705049266147,
      "eval_loss": 6.058621883392334,
      "eval_runtime": 765.0555,
      "eval_samples_per_second": 403.05,
      "eval_steps_per_second": 6.299,
      "step": 350000
    },
    {
      "epoch": 3.83,
      "learning_rate": 2.39480032162959e-06,
      "loss": 6.0609,
      "step": 350500
    },
    {
      "epoch": 3.83,
      "learning_rate": 2.3836326275350667e-06,
      "loss": 6.0572,
      "step": 351000
    },
    {
      "epoch": 3.84,
      "learning_rate": 2.3724649334405434e-06,
      "loss": 6.0571,
      "step": 351500
    },
    {
      "epoch": 3.85,
      "learning_rate": 2.36129723934602e-06,
      "loss": 6.0601,
      "step": 352000
    },
    {
      "epoch": 3.85,
      "learning_rate": 2.3501295452514964e-06,
      "loss": 6.0538,
      "step": 352500
    },
    {
      "epoch": 3.86,
      "learning_rate": 2.3389618511569736e-06,
      "loss": 6.0581,
      "step": 353000
    },
    {
      "epoch": 3.86,
      "learning_rate": 2.32779415706245e-06,
      "loss": 6.0567,
      "step": 353500
    },
    {
      "epoch": 3.87,
      "learning_rate": 2.3166264629679266e-06,
      "loss": 6.0472,
      "step": 354000
    },
    {
      "epoch": 3.87,
      "learning_rate": 2.3054587688734033e-06,
      "loss": 6.0571,
      "step": 354500
    },
    {
      "epoch": 3.88,
      "learning_rate": 2.29429107477888e-06,
      "loss": 6.0537,
      "step": 355000
    },
    {
      "epoch": 3.88,
      "learning_rate": 2.2831233806843564e-06,
      "loss": 6.0606,
      "step": 355500
    },
    {
      "epoch": 3.89,
      "learning_rate": 2.271955686589833e-06,
      "loss": 6.0597,
      "step": 356000
    },
    {
      "epoch": 3.89,
      "learning_rate": 2.26078799249531e-06,
      "loss": 6.0582,
      "step": 356500
    },
    {
      "epoch": 3.9,
      "learning_rate": 2.2496202984007865e-06,
      "loss": 6.0619,
      "step": 357000
    },
    {
      "epoch": 3.91,
      "learning_rate": 2.238452604306263e-06,
      "loss": 6.0546,
      "step": 357500
    },
    {
      "epoch": 3.91,
      "learning_rate": 2.2272849102117395e-06,
      "loss": 6.0532,
      "step": 358000
    },
    {
      "epoch": 3.92,
      "learning_rate": 2.2161172161172163e-06,
      "loss": 6.0623,
      "step": 358500
    },
    {
      "epoch": 3.92,
      "learning_rate": 2.204949522022693e-06,
      "loss": 6.0583,
      "step": 359000
    },
    {
      "epoch": 3.93,
      "learning_rate": 2.1937818279281693e-06,
      "loss": 6.0556,
      "step": 359500
    },
    {
      "epoch": 3.93,
      "learning_rate": 2.1826141338336464e-06,
      "loss": 6.0553,
      "step": 360000
    },
    {
      "epoch": 3.93,
      "eval_accuracy": 0.15059838625005148,
      "eval_loss": 6.058154582977295,
      "eval_runtime": 766.8207,
      "eval_samples_per_second": 402.123,
      "eval_steps_per_second": 6.284,
      "step": 360000
    },
    {
      "epoch": 3.94,
      "learning_rate": 2.1714464397391227e-06,
      "loss": 6.0556,
      "step": 360500
    },
    {
      "epoch": 3.94,
      "learning_rate": 2.1602787456445995e-06,
      "loss": 6.0597,
      "step": 361000
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.149111051550076e-06,
      "loss": 6.0592,
      "step": 361500
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.137943357455553e-06,
      "loss": 6.0553,
      "step": 362000
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.126775663361029e-06,
      "loss": 6.0618,
      "step": 362500
    },
    {
      "epoch": 3.97,
      "learning_rate": 2.115607969266506e-06,
      "loss": 6.0609,
      "step": 363000
    },
    {
      "epoch": 3.97,
      "learning_rate": 2.1044402751719826e-06,
      "loss": 6.0611,
      "step": 363500
    },
    {
      "epoch": 3.98,
      "learning_rate": 2.0932725810774594e-06,
      "loss": 6.0581,
      "step": 364000
    },
    {
      "epoch": 3.98,
      "learning_rate": 2.082104886982936e-06,
      "loss": 6.0582,
      "step": 364500
    },
    {
      "epoch": 3.99,
      "learning_rate": 2.0709371928884124e-06,
      "loss": 6.057,
      "step": 365000
    },
    {
      "epoch": 3.99,
      "learning_rate": 2.0597694987938895e-06,
      "loss": 6.0619,
      "step": 365500
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.048601804699366e-06,
      "loss": 6.0606,
      "step": 366000
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.0374341106048425e-06,
      "loss": 6.062,
      "step": 366500
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.0262664165103193e-06,
      "loss": 6.0607,
      "step": 367000
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.015098722415796e-06,
      "loss": 6.0568,
      "step": 367500
    },
    {
      "epoch": 4.02,
      "learning_rate": 2.0039310283212723e-06,
      "loss": 6.0627,
      "step": 368000
    },
    {
      "epoch": 4.03,
      "learning_rate": 1.992763334226749e-06,
      "loss": 6.0531,
      "step": 368500
    },
    {
      "epoch": 4.03,
      "learning_rate": 1.9815956401322257e-06,
      "loss": 6.0551,
      "step": 369000
    },
    {
      "epoch": 4.04,
      "learning_rate": 1.9704279460377025e-06,
      "loss": 6.0581,
      "step": 369500
    },
    {
      "epoch": 4.04,
      "learning_rate": 1.9592602519431788e-06,
      "loss": 6.053,
      "step": 370000
    },
    {
      "epoch": 4.04,
      "eval_accuracy": 0.15084313498363755,
      "eval_loss": 6.054418563842773,
      "eval_runtime": 768.4249,
      "eval_samples_per_second": 401.283,
      "eval_steps_per_second": 6.271,
      "step": 370000
    },
    {
      "epoch": 4.05,
      "learning_rate": 1.9480925578486555e-06,
      "loss": 6.0543,
      "step": 370500
    },
    {
      "epoch": 4.05,
      "learning_rate": 1.936924863754132e-06,
      "loss": 6.0564,
      "step": 371000
    },
    {
      "epoch": 4.06,
      "learning_rate": 1.925757169659609e-06,
      "loss": 6.0553,
      "step": 371500
    },
    {
      "epoch": 4.06,
      "learning_rate": 1.9145894755650852e-06,
      "loss": 6.0578,
      "step": 372000
    },
    {
      "epoch": 4.07,
      "learning_rate": 1.9034217814705621e-06,
      "loss": 6.0557,
      "step": 372500
    },
    {
      "epoch": 4.07,
      "learning_rate": 1.8922540873760387e-06,
      "loss": 6.057,
      "step": 373000
    },
    {
      "epoch": 4.08,
      "learning_rate": 1.8810863932815154e-06,
      "loss": 6.0585,
      "step": 373500
    },
    {
      "epoch": 4.09,
      "learning_rate": 1.8699186991869919e-06,
      "loss": 6.0568,
      "step": 374000
    },
    {
      "epoch": 4.09,
      "learning_rate": 1.8587510050924686e-06,
      "loss": 6.0545,
      "step": 374500
    },
    {
      "epoch": 4.1,
      "learning_rate": 1.8475833109979451e-06,
      "loss": 6.0505,
      "step": 375000
    },
    {
      "epoch": 4.1,
      "learning_rate": 1.836415616903422e-06,
      "loss": 6.0544,
      "step": 375500
    },
    {
      "epoch": 4.11,
      "learning_rate": 1.8252479228088984e-06,
      "loss": 6.051,
      "step": 376000
    },
    {
      "epoch": 4.11,
      "learning_rate": 1.8140802287143753e-06,
      "loss": 6.0632,
      "step": 376500
    },
    {
      "epoch": 4.12,
      "learning_rate": 1.8029125346198518e-06,
      "loss": 6.0551,
      "step": 377000
    },
    {
      "epoch": 4.12,
      "learning_rate": 1.7917448405253285e-06,
      "loss": 6.055,
      "step": 377500
    },
    {
      "epoch": 4.13,
      "learning_rate": 1.780577146430805e-06,
      "loss": 6.0542,
      "step": 378000
    },
    {
      "epoch": 4.13,
      "learning_rate": 1.7694094523362818e-06,
      "loss": 6.0566,
      "step": 378500
    },
    {
      "epoch": 4.14,
      "learning_rate": 1.7582417582417585e-06,
      "loss": 6.0562,
      "step": 379000
    },
    {
      "epoch": 4.15,
      "learning_rate": 1.747074064147235e-06,
      "loss": 6.0555,
      "step": 379500
    },
    {
      "epoch": 4.15,
      "learning_rate": 1.7359063700527117e-06,
      "loss": 6.0594,
      "step": 380000
    },
    {
      "epoch": 4.15,
      "eval_accuracy": 0.15068797496798347,
      "eval_loss": 6.05528450012207,
      "eval_runtime": 769.2453,
      "eval_samples_per_second": 400.855,
      "eval_steps_per_second": 6.265,
      "step": 380000
    },
    {
      "epoch": 4.16,
      "learning_rate": 1.7247386759581882e-06,
      "loss": 6.0476,
      "step": 380500
    },
    {
      "epoch": 4.16,
      "learning_rate": 1.7135709818636651e-06,
      "loss": 6.0547,
      "step": 381000
    },
    {
      "epoch": 4.17,
      "learning_rate": 1.7024032877691414e-06,
      "loss": 6.0561,
      "step": 381500
    },
    {
      "epoch": 4.17,
      "learning_rate": 1.6912355936746184e-06,
      "loss": 6.0536,
      "step": 382000
    },
    {
      "epoch": 4.18,
      "learning_rate": 1.6800678995800949e-06,
      "loss": 6.0622,
      "step": 382500
    },
    {
      "epoch": 4.18,
      "learning_rate": 1.6689002054855716e-06,
      "loss": 6.0516,
      "step": 383000
    },
    {
      "epoch": 4.19,
      "learning_rate": 1.6577325113910481e-06,
      "loss": 6.0615,
      "step": 383500
    },
    {
      "epoch": 4.19,
      "learning_rate": 1.6465648172965248e-06,
      "loss": 6.0608,
      "step": 384000
    },
    {
      "epoch": 4.2,
      "learning_rate": 1.6353971232020014e-06,
      "loss": 6.0583,
      "step": 384500
    },
    {
      "epoch": 4.21,
      "learning_rate": 1.624229429107478e-06,
      "loss": 6.0485,
      "step": 385000
    },
    {
      "epoch": 4.21,
      "learning_rate": 1.6130617350129546e-06,
      "loss": 6.0506,
      "step": 385500
    },
    {
      "epoch": 4.22,
      "learning_rate": 1.6018940409184313e-06,
      "loss": 6.0535,
      "step": 386000
    },
    {
      "epoch": 4.22,
      "learning_rate": 1.5907263468239078e-06,
      "loss": 6.0478,
      "step": 386500
    },
    {
      "epoch": 4.23,
      "learning_rate": 1.5795586527293845e-06,
      "loss": 6.0516,
      "step": 387000
    },
    {
      "epoch": 4.23,
      "learning_rate": 1.568390958634861e-06,
      "loss": 6.0516,
      "step": 387500
    },
    {
      "epoch": 4.24,
      "learning_rate": 1.557223264540338e-06,
      "loss": 6.0541,
      "step": 388000
    },
    {
      "epoch": 4.24,
      "learning_rate": 1.5460555704458145e-06,
      "loss": 6.0512,
      "step": 388500
    },
    {
      "epoch": 4.25,
      "learning_rate": 1.5348878763512912e-06,
      "loss": 6.0507,
      "step": 389000
    },
    {
      "epoch": 4.25,
      "learning_rate": 1.5237201822567677e-06,
      "loss": 6.0526,
      "step": 389500
    },
    {
      "epoch": 4.26,
      "learning_rate": 1.5125524881622444e-06,
      "loss": 6.0488,
      "step": 390000
    },
    {
      "epoch": 4.26,
      "eval_accuracy": 0.15088102471328585,
      "eval_loss": 6.052678108215332,
      "eval_runtime": 767.5405,
      "eval_samples_per_second": 401.746,
      "eval_steps_per_second": 6.278,
      "step": 390000
    },
    {
      "epoch": 4.27,
      "learning_rate": 1.501384794067721e-06,
      "loss": 6.0576,
      "step": 390500
    },
    {
      "epoch": 4.27,
      "learning_rate": 1.4902170999731977e-06,
      "loss": 6.0522,
      "step": 391000
    },
    {
      "epoch": 4.28,
      "learning_rate": 1.4790494058786742e-06,
      "loss": 6.0536,
      "step": 391500
    },
    {
      "epoch": 4.28,
      "learning_rate": 1.467881711784151e-06,
      "loss": 6.0523,
      "step": 392000
    },
    {
      "epoch": 4.29,
      "learning_rate": 1.4567140176896274e-06,
      "loss": 6.0547,
      "step": 392500
    },
    {
      "epoch": 4.29,
      "learning_rate": 1.4455463235951041e-06,
      "loss": 6.0515,
      "step": 393000
    },
    {
      "epoch": 4.3,
      "learning_rate": 1.4343786295005807e-06,
      "loss": 6.0527,
      "step": 393500
    },
    {
      "epoch": 4.3,
      "learning_rate": 1.4232109354060574e-06,
      "loss": 6.0545,
      "step": 394000
    },
    {
      "epoch": 4.31,
      "learning_rate": 1.4120432413115339e-06,
      "loss": 6.0508,
      "step": 394500
    },
    {
      "epoch": 4.31,
      "learning_rate": 1.4008755472170108e-06,
      "loss": 6.0538,
      "step": 395000
    },
    {
      "epoch": 4.32,
      "learning_rate": 1.3897078531224875e-06,
      "loss": 6.0537,
      "step": 395500
    },
    {
      "epoch": 4.33,
      "learning_rate": 1.378540159027964e-06,
      "loss": 6.0529,
      "step": 396000
    },
    {
      "epoch": 4.33,
      "learning_rate": 1.3673724649334408e-06,
      "loss": 6.0535,
      "step": 396500
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.3562047708389173e-06,
      "loss": 6.0488,
      "step": 397000
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.345037076744394e-06,
      "loss": 6.0511,
      "step": 397500
    },
    {
      "epoch": 4.35,
      "learning_rate": 1.3338693826498705e-06,
      "loss": 6.0541,
      "step": 398000
    },
    {
      "epoch": 4.35,
      "learning_rate": 1.3227016885553472e-06,
      "loss": 6.0524,
      "step": 398500
    },
    {
      "epoch": 4.36,
      "learning_rate": 1.3115339944608238e-06,
      "loss": 6.0529,
      "step": 399000
    },
    {
      "epoch": 4.36,
      "learning_rate": 1.3003663003663005e-06,
      "loss": 6.0483,
      "step": 399500
    },
    {
      "epoch": 4.37,
      "learning_rate": 1.289198606271777e-06,
      "loss": 6.051,
      "step": 400000
    },
    {
      "epoch": 4.37,
      "eval_accuracy": 0.15087456143454891,
      "eval_loss": 6.0515947341918945,
      "eval_runtime": 767.8289,
      "eval_samples_per_second": 401.595,
      "eval_steps_per_second": 6.276,
      "step": 400000
    },
    {
      "epoch": 4.37,
      "learning_rate": 1.278030912177254e-06,
      "loss": 6.0537,
      "step": 400500
    },
    {
      "epoch": 4.38,
      "learning_rate": 1.2668632180827304e-06,
      "loss": 6.0557,
      "step": 401000
    },
    {
      "epoch": 4.39,
      "learning_rate": 1.2556955239882071e-06,
      "loss": 6.0521,
      "step": 401500
    },
    {
      "epoch": 4.39,
      "learning_rate": 1.2445278298936837e-06,
      "loss": 6.0551,
      "step": 402000
    },
    {
      "epoch": 4.4,
      "learning_rate": 1.2333601357991602e-06,
      "loss": 6.052,
      "step": 402500
    },
    {
      "epoch": 4.4,
      "learning_rate": 1.2221924417046369e-06,
      "loss": 6.0519,
      "step": 403000
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.2110247476101136e-06,
      "loss": 6.0519,
      "step": 403500
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.1998570535155903e-06,
      "loss": 6.0488,
      "step": 404000
    },
    {
      "epoch": 4.42,
      "learning_rate": 1.1886893594210668e-06,
      "loss": 6.0512,
      "step": 404500
    },
    {
      "epoch": 4.42,
      "learning_rate": 1.1775216653265436e-06,
      "loss": 6.0553,
      "step": 405000
    },
    {
      "epoch": 4.43,
      "learning_rate": 1.16635397123202e-06,
      "loss": 6.0468,
      "step": 405500
    },
    {
      "epoch": 4.44,
      "learning_rate": 1.1551862771374968e-06,
      "loss": 6.048,
      "step": 406000
    },
    {
      "epoch": 4.44,
      "learning_rate": 1.1440185830429735e-06,
      "loss": 6.0517,
      "step": 406500
    },
    {
      "epoch": 4.45,
      "learning_rate": 1.13285088894845e-06,
      "loss": 6.0532,
      "step": 407000
    },
    {
      "epoch": 4.45,
      "learning_rate": 1.1216831948539268e-06,
      "loss": 6.0515,
      "step": 407500
    },
    {
      "epoch": 4.46,
      "learning_rate": 1.1105155007594033e-06,
      "loss": 6.0566,
      "step": 408000
    },
    {
      "epoch": 4.46,
      "learning_rate": 1.09934780666488e-06,
      "loss": 6.0526,
      "step": 408500
    },
    {
      "epoch": 4.47,
      "learning_rate": 1.0881801125703565e-06,
      "loss": 6.0512,
      "step": 409000
    },
    {
      "epoch": 4.47,
      "learning_rate": 1.0770124184758332e-06,
      "loss": 6.0525,
      "step": 409500
    },
    {
      "epoch": 4.48,
      "learning_rate": 1.06584472438131e-06,
      "loss": 6.0553,
      "step": 410000
    },
    {
      "epoch": 4.48,
      "eval_accuracy": 0.15085937486796927,
      "eval_loss": 6.051754951477051,
      "eval_runtime": 767.6885,
      "eval_samples_per_second": 401.668,
      "eval_steps_per_second": 6.277,
      "step": 410000
    },
    {
      "epoch": 4.48,
      "learning_rate": 1.0546770302867864e-06,
      "loss": 6.0525,
      "step": 410500
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.0435093361922632e-06,
      "loss": 6.046,
      "step": 411000
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.0323416420977397e-06,
      "loss": 6.0476,
      "step": 411500
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.0211739480032164e-06,
      "loss": 6.0523,
      "step": 412000
    },
    {
      "epoch": 4.51,
      "learning_rate": 1.010006253908693e-06,
      "loss": 6.0543,
      "step": 412500
    },
    {
      "epoch": 4.51,
      "learning_rate": 9.988385598141696e-07,
      "loss": 6.0544,
      "step": 413000
    },
    {
      "epoch": 4.52,
      "learning_rate": 9.876708657196464e-07,
      "loss": 6.0476,
      "step": 413500
    },
    {
      "epoch": 4.52,
      "learning_rate": 9.765031716251229e-07,
      "loss": 6.0489,
      "step": 414000
    },
    {
      "epoch": 4.53,
      "learning_rate": 9.653354775305996e-07,
      "loss": 6.0556,
      "step": 414500
    },
    {
      "epoch": 4.53,
      "learning_rate": 9.54167783436076e-07,
      "loss": 6.0573,
      "step": 415000
    },
    {
      "epoch": 4.54,
      "learning_rate": 9.430000893415528e-07,
      "loss": 6.0534,
      "step": 415500
    },
    {
      "epoch": 4.54,
      "learning_rate": 9.318323952470294e-07,
      "loss": 6.0534,
      "step": 416000
    },
    {
      "epoch": 4.55,
      "learning_rate": 9.20664701152506e-07,
      "loss": 6.055,
      "step": 416500
    },
    {
      "epoch": 4.56,
      "learning_rate": 9.094970070579827e-07,
      "loss": 6.0508,
      "step": 417000
    },
    {
      "epoch": 4.56,
      "learning_rate": 8.983293129634593e-07,
      "loss": 6.049,
      "step": 417500
    },
    {
      "epoch": 4.57,
      "learning_rate": 8.87161618868936e-07,
      "loss": 6.0533,
      "step": 418000
    },
    {
      "epoch": 4.57,
      "learning_rate": 8.759939247744126e-07,
      "loss": 6.0532,
      "step": 418500
    },
    {
      "epoch": 4.58,
      "learning_rate": 8.648262306798893e-07,
      "loss": 6.0564,
      "step": 419000
    },
    {
      "epoch": 4.58,
      "learning_rate": 8.53658536585366e-07,
      "loss": 6.0456,
      "step": 419500
    },
    {
      "epoch": 4.59,
      "learning_rate": 8.424908424908426e-07,
      "loss": 6.0507,
      "step": 420000
    },
    {
      "epoch": 4.59,
      "eval_accuracy": 0.15092157875145593,
      "eval_loss": 6.051989555358887,
      "eval_runtime": 766.02,
      "eval_samples_per_second": 402.543,
      "eval_steps_per_second": 6.291,
      "step": 420000
    },
    {
      "epoch": 4.59,
      "learning_rate": 8.313231483963193e-07,
      "loss": 6.0552,
      "step": 420500
    },
    {
      "epoch": 4.6,
      "learning_rate": 8.201554543017959e-07,
      "loss": 6.0466,
      "step": 421000
    },
    {
      "epoch": 4.6,
      "learning_rate": 8.089877602072725e-07,
      "loss": 6.0498,
      "step": 421500
    },
    {
      "epoch": 4.61,
      "learning_rate": 7.978200661127491e-07,
      "loss": 6.0534,
      "step": 422000
    },
    {
      "epoch": 4.62,
      "learning_rate": 7.866523720182258e-07,
      "loss": 6.0522,
      "step": 422500
    },
    {
      "epoch": 4.62,
      "learning_rate": 7.754846779237024e-07,
      "loss": 6.0505,
      "step": 423000
    },
    {
      "epoch": 4.63,
      "learning_rate": 7.64316983829179e-07,
      "loss": 6.0507,
      "step": 423500
    },
    {
      "epoch": 4.63,
      "learning_rate": 7.531492897346557e-07,
      "loss": 6.0505,
      "step": 424000
    },
    {
      "epoch": 4.64,
      "learning_rate": 7.419815956401323e-07,
      "loss": 6.0548,
      "step": 424500
    },
    {
      "epoch": 4.64,
      "learning_rate": 7.30813901545609e-07,
      "loss": 6.0519,
      "step": 425000
    },
    {
      "epoch": 4.65,
      "learning_rate": 7.196462074510856e-07,
      "loss": 6.051,
      "step": 425500
    },
    {
      "epoch": 4.65,
      "learning_rate": 7.084785133565622e-07,
      "loss": 6.0471,
      "step": 426000
    },
    {
      "epoch": 4.66,
      "learning_rate": 6.973108192620388e-07,
      "loss": 6.0494,
      "step": 426500
    },
    {
      "epoch": 4.66,
      "learning_rate": 6.861431251675155e-07,
      "loss": 6.056,
      "step": 427000
    },
    {
      "epoch": 4.67,
      "learning_rate": 6.749754310729921e-07,
      "loss": 6.0462,
      "step": 427500
    },
    {
      "epoch": 4.68,
      "learning_rate": 6.638077369784687e-07,
      "loss": 6.0505,
      "step": 428000
    },
    {
      "epoch": 4.68,
      "learning_rate": 6.526400428839454e-07,
      "loss": 6.0491,
      "step": 428500
    },
    {
      "epoch": 4.69,
      "learning_rate": 6.41472348789422e-07,
      "loss": 6.0561,
      "step": 429000
    },
    {
      "epoch": 4.69,
      "learning_rate": 6.303046546948986e-07,
      "loss": 6.0471,
      "step": 429500
    },
    {
      "epoch": 4.7,
      "learning_rate": 6.191369606003752e-07,
      "loss": 6.0514,
      "step": 430000
    },
    {
      "epoch": 4.7,
      "eval_accuracy": 0.15087106042233073,
      "eval_loss": 6.050050258636475,
      "eval_runtime": 765.9306,
      "eval_samples_per_second": 402.59,
      "eval_steps_per_second": 6.292,
      "step": 430000
    },
    {
      "epoch": 4.7,
      "learning_rate": 6.079692665058519e-07,
      "loss": 6.0518,
      "step": 430500
    },
    {
      "epoch": 4.71,
      "learning_rate": 5.968015724113286e-07,
      "loss": 6.0492,
      "step": 431000
    },
    {
      "epoch": 4.71,
      "learning_rate": 5.856338783168053e-07,
      "loss": 6.0491,
      "step": 431500
    },
    {
      "epoch": 4.72,
      "learning_rate": 5.744661842222819e-07,
      "loss": 6.0491,
      "step": 432000
    },
    {
      "epoch": 4.72,
      "learning_rate": 5.632984901277585e-07,
      "loss": 6.0483,
      "step": 432500
    },
    {
      "epoch": 4.73,
      "learning_rate": 5.521307960332351e-07,
      "loss": 6.0511,
      "step": 433000
    },
    {
      "epoch": 4.74,
      "learning_rate": 5.409631019387117e-07,
      "loss": 6.0449,
      "step": 433500
    },
    {
      "epoch": 4.74,
      "learning_rate": 5.297954078441884e-07,
      "loss": 6.0506,
      "step": 434000
    },
    {
      "epoch": 4.75,
      "learning_rate": 5.18627713749665e-07,
      "loss": 6.0518,
      "step": 434500
    },
    {
      "epoch": 4.75,
      "learning_rate": 5.074600196551417e-07,
      "loss": 6.0542,
      "step": 435000
    },
    {
      "epoch": 4.76,
      "learning_rate": 4.962923255606183e-07,
      "loss": 6.0497,
      "step": 435500
    },
    {
      "epoch": 4.76,
      "learning_rate": 4.851246314660949e-07,
      "loss": 6.0566,
      "step": 436000
    },
    {
      "epoch": 4.77,
      "learning_rate": 4.7395693737157154e-07,
      "loss": 6.048,
      "step": 436500
    },
    {
      "epoch": 4.77,
      "learning_rate": 4.6278924327704816e-07,
      "loss": 6.0521,
      "step": 437000
    },
    {
      "epoch": 4.78,
      "learning_rate": 4.516215491825248e-07,
      "loss": 6.0499,
      "step": 437500
    },
    {
      "epoch": 4.78,
      "learning_rate": 4.4045385508800144e-07,
      "loss": 6.0444,
      "step": 438000
    },
    {
      "epoch": 4.79,
      "learning_rate": 4.2928616099347806e-07,
      "loss": 6.0455,
      "step": 438500
    },
    {
      "epoch": 4.8,
      "learning_rate": 4.181184668989548e-07,
      "loss": 6.0456,
      "step": 439000
    },
    {
      "epoch": 4.8,
      "learning_rate": 4.069507728044314e-07,
      "loss": 6.0493,
      "step": 439500
    },
    {
      "epoch": 4.81,
      "learning_rate": 3.95783078709908e-07,
      "loss": 6.0511,
      "step": 440000
    },
    {
      "epoch": 4.81,
      "eval_accuracy": 0.15107602014161378,
      "eval_loss": 6.049585342407227,
      "eval_runtime": 766.2027,
      "eval_samples_per_second": 402.447,
      "eval_steps_per_second": 6.289,
      "step": 440000
    },
    {
      "epoch": 4.81,
      "learning_rate": 3.846153846153847e-07,
      "loss": 6.0492,
      "step": 440500
    },
    {
      "epoch": 4.82,
      "learning_rate": 3.734476905208613e-07,
      "loss": 6.0431,
      "step": 441000
    },
    {
      "epoch": 4.82,
      "learning_rate": 3.622799964263379e-07,
      "loss": 6.056,
      "step": 441500
    },
    {
      "epoch": 4.83,
      "learning_rate": 3.511123023318146e-07,
      "loss": 6.0513,
      "step": 442000
    },
    {
      "epoch": 4.83,
      "learning_rate": 3.399446082372912e-07,
      "loss": 6.0551,
      "step": 442500
    },
    {
      "epoch": 4.84,
      "learning_rate": 3.287769141427678e-07,
      "loss": 6.0482,
      "step": 443000
    },
    {
      "epoch": 4.84,
      "learning_rate": 3.176092200482445e-07,
      "loss": 6.0516,
      "step": 443500
    },
    {
      "epoch": 4.85,
      "learning_rate": 3.064415259537211e-07,
      "loss": 6.0462,
      "step": 444000
    },
    {
      "epoch": 4.86,
      "learning_rate": 2.952738318591977e-07,
      "loss": 6.0505,
      "step": 444500
    },
    {
      "epoch": 4.86,
      "learning_rate": 2.841061377646744e-07,
      "loss": 6.0497,
      "step": 445000
    },
    {
      "epoch": 4.87,
      "learning_rate": 2.72938443670151e-07,
      "loss": 6.0456,
      "step": 445500
    },
    {
      "epoch": 4.87,
      "learning_rate": 2.6177074957562767e-07,
      "loss": 6.0492,
      "step": 446000
    },
    {
      "epoch": 4.88,
      "learning_rate": 2.506030554811043e-07,
      "loss": 6.0479,
      "step": 446500
    },
    {
      "epoch": 4.88,
      "learning_rate": 2.394353613865809e-07,
      "loss": 6.0468,
      "step": 447000
    },
    {
      "epoch": 4.89,
      "learning_rate": 2.2826766729205754e-07,
      "loss": 6.0508,
      "step": 447500
    },
    {
      "epoch": 4.89,
      "learning_rate": 2.1709997319753418e-07,
      "loss": 6.0468,
      "step": 448000
    },
    {
      "epoch": 4.9,
      "learning_rate": 2.0593227910301083e-07,
      "loss": 6.0459,
      "step": 448500
    },
    {
      "epoch": 4.9,
      "learning_rate": 1.9476458500848747e-07,
      "loss": 6.0509,
      "step": 449000
    },
    {
      "epoch": 4.91,
      "learning_rate": 1.835968909139641e-07,
      "loss": 6.0468,
      "step": 449500
    },
    {
      "epoch": 4.92,
      "learning_rate": 1.7242919681944075e-07,
      "loss": 6.0527,
      "step": 450000
    },
    {
      "epoch": 4.92,
      "eval_accuracy": 0.15092618811002353,
      "eval_loss": 6.0493364334106445,
      "eval_runtime": 766.4167,
      "eval_samples_per_second": 402.335,
      "eval_steps_per_second": 6.288,
      "step": 450000
    },
    {
      "epoch": 4.92,
      "learning_rate": 1.6126150272491737e-07,
      "loss": 6.0532,
      "step": 450500
    },
    {
      "epoch": 4.93,
      "learning_rate": 1.50093808630394e-07,
      "loss": 6.0454,
      "step": 451000
    },
    {
      "epoch": 4.93,
      "learning_rate": 1.3892611453587063e-07,
      "loss": 6.0459,
      "step": 451500
    },
    {
      "epoch": 4.94,
      "learning_rate": 1.277584204413473e-07,
      "loss": 6.0514,
      "step": 452000
    },
    {
      "epoch": 4.94,
      "learning_rate": 1.1659072634682393e-07,
      "loss": 6.0524,
      "step": 452500
    },
    {
      "epoch": 4.95,
      "learning_rate": 1.0542303225230056e-07,
      "loss": 6.0488,
      "step": 453000
    },
    {
      "epoch": 4.95,
      "learning_rate": 9.425533815777719e-08,
      "loss": 6.0529,
      "step": 453500
    },
    {
      "epoch": 4.96,
      "learning_rate": 8.308764406325384e-08,
      "loss": 6.0435,
      "step": 454000
    },
    {
      "epoch": 4.96,
      "learning_rate": 7.191994996873047e-08,
      "loss": 6.0487,
      "step": 454500
    },
    {
      "epoch": 4.97,
      "learning_rate": 6.07522558742071e-08,
      "loss": 6.0485,
      "step": 455000
    },
    {
      "epoch": 4.98,
      "learning_rate": 4.958456177968374e-08,
      "loss": 6.0532,
      "step": 455500
    },
    {
      "epoch": 4.98,
      "learning_rate": 3.841686768516037e-08,
      "loss": 6.0425,
      "step": 456000
    },
    {
      "epoch": 4.99,
      "learning_rate": 2.7249173590637007e-08,
      "loss": 6.0476,
      "step": 456500
    },
    {
      "epoch": 4.99,
      "learning_rate": 1.6081479496113643e-08,
      "loss": 6.0489,
      "step": 457000
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.91378540159028e-09,
      "loss": 6.0543,
      "step": 457500
    },
    {
      "epoch": 5.0,
      "step": 457720,
      "total_flos": 8.098601682117591e+17,
      "train_loss": 6.211948149485322,
      "train_runtime": 192213.636,
      "train_samples_per_second": 152.402,
      "train_steps_per_second": 2.381
    }
  ],
  "logging_steps": 500,
  "max_steps": 457720,
  "num_train_epochs": 5,
  "save_steps": 10000,
  "total_flos": 8.098601682117591e+17,
  "trial_name": null,
  "trial_params": null
}