{
  "best_metric": 6.212860107421875,
  "best_model_checkpoint": "HBERTv1_emb_compress_48_L10_H128_A2/checkpoint-350000",
  "epoch": 5.0,
  "eval_steps": 10000,
  "global_step": 366175,
  "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": 10.2904,
      "step": 500
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 10.2232,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.5e-06,
      "loss": 10.1752,
      "step": 1500
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 10.1161,
      "step": 2000
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.5e-06,
      "loss": 10.0425,
      "step": 2500
    },
    {
      "epoch": 0.04,
      "learning_rate": 3e-06,
      "loss": 9.9545,
      "step": 3000
    },
    {
      "epoch": 0.05,
      "learning_rate": 3.5e-06,
      "loss": 9.8497,
      "step": 3500
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.000000000000001e-06,
      "loss": 9.7291,
      "step": 4000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.5e-06,
      "loss": 9.5934,
      "step": 4500
    },
    {
      "epoch": 0.07,
      "learning_rate": 5e-06,
      "loss": 9.4429,
      "step": 5000
    },
    {
      "epoch": 0.08,
      "learning_rate": 5.500000000000001e-06,
      "loss": 9.2824,
      "step": 5500
    },
    {
      "epoch": 0.08,
      "learning_rate": 6e-06,
      "loss": 9.1117,
      "step": 6000
    },
    {
      "epoch": 0.09,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 8.9342,
      "step": 6500
    },
    {
      "epoch": 0.1,
      "learning_rate": 7e-06,
      "loss": 8.7503,
      "step": 7000
    },
    {
      "epoch": 0.1,
      "learning_rate": 7.500000000000001e-06,
      "loss": 8.5662,
      "step": 7500
    },
    {
      "epoch": 0.11,
      "learning_rate": 8.000000000000001e-06,
      "loss": 8.3825,
      "step": 8000
    },
    {
      "epoch": 0.12,
      "learning_rate": 8.5e-06,
      "loss": 8.2021,
      "step": 8500
    },
    {
      "epoch": 0.12,
      "learning_rate": 9e-06,
      "loss": 8.0279,
      "step": 9000
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.5e-06,
      "loss": 7.8638,
      "step": 9500
    },
    {
      "epoch": 0.14,
      "learning_rate": 1e-05,
      "loss": 7.7198,
      "step": 10000
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 0.06362994354363025,
      "eval_loss": 7.64456844329834,
      "eval_runtime": 592.6651,
      "eval_samples_per_second": 520.287,
      "eval_steps_per_second": 6.505,
      "step": 10000
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.98596195690321e-06,
      "loss": 7.5854,
      "step": 10500
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.971923913806416e-06,
      "loss": 7.4734,
      "step": 11000
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.957885870709624e-06,
      "loss": 7.3871,
      "step": 11500
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.943847827612831e-06,
      "loss": 7.3185,
      "step": 12000
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.92980978451604e-06,
      "loss": 7.2718,
      "step": 12500
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.915771741419246e-06,
      "loss": 7.2275,
      "step": 13000
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.901733698322455e-06,
      "loss": 7.2023,
      "step": 13500
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.887695655225662e-06,
      "loss": 7.173,
      "step": 14000
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.87365761212887e-06,
      "loss": 7.1471,
      "step": 14500
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.859619569032077e-06,
      "loss": 7.119,
      "step": 15000
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.845581525935285e-06,
      "loss": 7.1001,
      "step": 15500
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.831543482838492e-06,
      "loss": 7.0771,
      "step": 16000
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.8175054397417e-06,
      "loss": 7.0519,
      "step": 16500
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.803467396644909e-06,
      "loss": 7.0345,
      "step": 17000
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.789429353548117e-06,
      "loss": 7.0182,
      "step": 17500
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.775391310451324e-06,
      "loss": 7.0053,
      "step": 18000
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.761353267354533e-06,
      "loss": 6.9873,
      "step": 18500
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.74731522425774e-06,
      "loss": 6.9777,
      "step": 19000
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.733277181160948e-06,
      "loss": 6.9601,
      "step": 19500
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.719239138064155e-06,
      "loss": 6.9424,
      "step": 20000
    },
    {
      "epoch": 0.27,
      "eval_accuracy": 0.08250771948999101,
      "eval_loss": 6.937928676605225,
      "eval_runtime": 593.2654,
      "eval_samples_per_second": 519.761,
      "eval_steps_per_second": 6.498,
      "step": 20000
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.705201094967363e-06,
      "loss": 6.9304,
      "step": 20500
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.69116305187057e-06,
      "loss": 6.9208,
      "step": 21000
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.677125008773778e-06,
      "loss": 6.9118,
      "step": 21500
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.663086965676985e-06,
      "loss": 6.8899,
      "step": 22000
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.649048922580194e-06,
      "loss": 6.8851,
      "step": 22500
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.6350108794834e-06,
      "loss": 6.8754,
      "step": 23000
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.620972836386609e-06,
      "loss": 6.8624,
      "step": 23500
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.606934793289815e-06,
      "loss": 6.847,
      "step": 24000
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.592896750193024e-06,
      "loss": 6.8442,
      "step": 24500
    },
    {
      "epoch": 0.34,
      "learning_rate": 9.57885870709623e-06,
      "loss": 6.834,
      "step": 25000
    },
    {
      "epoch": 0.35,
      "learning_rate": 9.564820663999439e-06,
      "loss": 6.8277,
      "step": 25500
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.550782620902646e-06,
      "loss": 6.8179,
      "step": 26000
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.536744577805854e-06,
      "loss": 6.8111,
      "step": 26500
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.522706534709061e-06,
      "loss": 6.7985,
      "step": 27000
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.50866849161227e-06,
      "loss": 6.7856,
      "step": 27500
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.494630448515478e-06,
      "loss": 6.7789,
      "step": 28000
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.480592405418685e-06,
      "loss": 6.7815,
      "step": 28500
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.466554362321893e-06,
      "loss": 6.7675,
      "step": 29000
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.452516319225102e-06,
      "loss": 6.7582,
      "step": 29500
    },
    {
      "epoch": 0.41,
      "learning_rate": 9.438478276128308e-06,
      "loss": 6.7559,
      "step": 30000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.09742729395013566,
      "eval_loss": 6.749167442321777,
      "eval_runtime": 591.9968,
      "eval_samples_per_second": 520.874,
      "eval_steps_per_second": 6.512,
      "step": 30000
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.424440233031517e-06,
      "loss": 6.7479,
      "step": 30500
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.410402189934724e-06,
      "loss": 6.7427,
      "step": 31000
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.396364146837932e-06,
      "loss": 6.7306,
      "step": 31500
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.382326103741139e-06,
      "loss": 6.7271,
      "step": 32000
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.368288060644347e-06,
      "loss": 6.7232,
      "step": 32500
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.354250017547554e-06,
      "loss": 6.7147,
      "step": 33000
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.340211974450763e-06,
      "loss": 6.7085,
      "step": 33500
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.32617393135397e-06,
      "loss": 6.7078,
      "step": 34000
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.312135888257178e-06,
      "loss": 6.7029,
      "step": 34500
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.298097845160386e-06,
      "loss": 6.6892,
      "step": 35000
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.284059802063593e-06,
      "loss": 6.6938,
      "step": 35500
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.270021758966801e-06,
      "loss": 6.6861,
      "step": 36000
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.255983715870008e-06,
      "loss": 6.6832,
      "step": 36500
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.241945672773217e-06,
      "loss": 6.6733,
      "step": 37000
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.227907629676423e-06,
      "loss": 6.6687,
      "step": 37500
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.213869586579632e-06,
      "loss": 6.6656,
      "step": 38000
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.199831543482839e-06,
      "loss": 6.6596,
      "step": 38500
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.185793500386047e-06,
      "loss": 6.6591,
      "step": 39000
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.171755457289254e-06,
      "loss": 6.6593,
      "step": 39500
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.157717414192462e-06,
      "loss": 6.6513,
      "step": 40000
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.10470041728507712,
      "eval_loss": 6.645530700683594,
      "eval_runtime": 590.5895,
      "eval_samples_per_second": 522.116,
      "eval_steps_per_second": 6.527,
      "step": 40000
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.143679371095669e-06,
      "loss": 6.6443,
      "step": 40500
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.129641327998877e-06,
      "loss": 6.636,
      "step": 41000
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.115603284902086e-06,
      "loss": 6.6321,
      "step": 41500
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.101565241805294e-06,
      "loss": 6.6254,
      "step": 42000
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.087527198708501e-06,
      "loss": 6.6302,
      "step": 42500
    },
    {
      "epoch": 0.59,
      "learning_rate": 9.07348915561171e-06,
      "loss": 6.6243,
      "step": 43000
    },
    {
      "epoch": 0.59,
      "learning_rate": 9.059451112514916e-06,
      "loss": 6.6213,
      "step": 43500
    },
    {
      "epoch": 0.6,
      "learning_rate": 9.045413069418125e-06,
      "loss": 6.6154,
      "step": 44000
    },
    {
      "epoch": 0.61,
      "learning_rate": 9.031375026321332e-06,
      "loss": 6.6123,
      "step": 44500
    },
    {
      "epoch": 0.61,
      "learning_rate": 9.01733698322454e-06,
      "loss": 6.6047,
      "step": 45000
    },
    {
      "epoch": 0.62,
      "learning_rate": 9.003298940127747e-06,
      "loss": 6.6071,
      "step": 45500
    },
    {
      "epoch": 0.63,
      "learning_rate": 8.989260897030955e-06,
      "loss": 6.605,
      "step": 46000
    },
    {
      "epoch": 0.63,
      "learning_rate": 8.975222853934162e-06,
      "loss": 6.602,
      "step": 46500
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.96118481083737e-06,
      "loss": 6.5942,
      "step": 47000
    },
    {
      "epoch": 0.65,
      "learning_rate": 8.947146767740577e-06,
      "loss": 6.5868,
      "step": 47500
    },
    {
      "epoch": 0.66,
      "learning_rate": 8.933108724643786e-06,
      "loss": 6.5934,
      "step": 48000
    },
    {
      "epoch": 0.66,
      "learning_rate": 8.919070681546992e-06,
      "loss": 6.5847,
      "step": 48500
    },
    {
      "epoch": 0.67,
      "learning_rate": 8.905032638450201e-06,
      "loss": 6.5833,
      "step": 49000
    },
    {
      "epoch": 0.68,
      "learning_rate": 8.890994595353408e-06,
      "loss": 6.5809,
      "step": 49500
    },
    {
      "epoch": 0.68,
      "learning_rate": 8.876956552256616e-06,
      "loss": 6.5706,
      "step": 50000
    },
    {
      "epoch": 0.68,
      "eval_accuracy": 0.11129192039684155,
      "eval_loss": 6.573412895202637,
      "eval_runtime": 591.297,
      "eval_samples_per_second": 521.491,
      "eval_steps_per_second": 6.52,
      "step": 50000
    },
    {
      "epoch": 0.69,
      "learning_rate": 8.862918509159823e-06,
      "loss": 6.5695,
      "step": 50500
    },
    {
      "epoch": 0.7,
      "learning_rate": 8.848880466063031e-06,
      "loss": 6.5655,
      "step": 51000
    },
    {
      "epoch": 0.7,
      "learning_rate": 8.834842422966238e-06,
      "loss": 6.5704,
      "step": 51500
    },
    {
      "epoch": 0.71,
      "learning_rate": 8.820804379869447e-06,
      "loss": 6.5545,
      "step": 52000
    },
    {
      "epoch": 0.72,
      "learning_rate": 8.806766336772655e-06,
      "loss": 6.5611,
      "step": 52500
    },
    {
      "epoch": 0.72,
      "learning_rate": 8.792728293675862e-06,
      "loss": 6.5557,
      "step": 53000
    },
    {
      "epoch": 0.73,
      "learning_rate": 8.77869025057907e-06,
      "loss": 6.5536,
      "step": 53500
    },
    {
      "epoch": 0.74,
      "learning_rate": 8.764652207482279e-06,
      "loss": 6.5473,
      "step": 54000
    },
    {
      "epoch": 0.74,
      "learning_rate": 8.750614164385485e-06,
      "loss": 6.5497,
      "step": 54500
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.736576121288694e-06,
      "loss": 6.5457,
      "step": 55000
    },
    {
      "epoch": 0.76,
      "learning_rate": 8.7225380781919e-06,
      "loss": 6.5389,
      "step": 55500
    },
    {
      "epoch": 0.76,
      "learning_rate": 8.708500035095109e-06,
      "loss": 6.5383,
      "step": 56000
    },
    {
      "epoch": 0.77,
      "learning_rate": 8.694461991998316e-06,
      "loss": 6.5412,
      "step": 56500
    },
    {
      "epoch": 0.78,
      "learning_rate": 8.680423948901524e-06,
      "loss": 6.5289,
      "step": 57000
    },
    {
      "epoch": 0.79,
      "learning_rate": 8.666385905804731e-06,
      "loss": 6.5364,
      "step": 57500
    },
    {
      "epoch": 0.79,
      "learning_rate": 8.65234786270794e-06,
      "loss": 6.528,
      "step": 58000
    },
    {
      "epoch": 0.8,
      "learning_rate": 8.638309819611146e-06,
      "loss": 6.5217,
      "step": 58500
    },
    {
      "epoch": 0.81,
      "learning_rate": 8.624271776514355e-06,
      "loss": 6.5233,
      "step": 59000
    },
    {
      "epoch": 0.81,
      "learning_rate": 8.610233733417563e-06,
      "loss": 6.5266,
      "step": 59500
    },
    {
      "epoch": 0.82,
      "learning_rate": 8.59619569032077e-06,
      "loss": 6.5236,
      "step": 60000
    },
    {
      "epoch": 0.82,
      "eval_accuracy": 0.11700887135181261,
      "eval_loss": 6.51757287979126,
      "eval_runtime": 593.5031,
      "eval_samples_per_second": 519.552,
      "eval_steps_per_second": 6.495,
      "step": 60000
    },
    {
      "epoch": 0.83,
      "learning_rate": 8.582157647223978e-06,
      "loss": 6.5112,
      "step": 60500
    },
    {
      "epoch": 0.83,
      "learning_rate": 8.568119604127185e-06,
      "loss": 6.5119,
      "step": 61000
    },
    {
      "epoch": 0.84,
      "learning_rate": 8.554081561030394e-06,
      "loss": 6.511,
      "step": 61500
    },
    {
      "epoch": 0.85,
      "learning_rate": 8.5400435179336e-06,
      "loss": 6.5091,
      "step": 62000
    },
    {
      "epoch": 0.85,
      "learning_rate": 8.526005474836809e-06,
      "loss": 6.5075,
      "step": 62500
    },
    {
      "epoch": 0.86,
      "learning_rate": 8.511967431740016e-06,
      "loss": 6.5036,
      "step": 63000
    },
    {
      "epoch": 0.87,
      "learning_rate": 8.497929388643224e-06,
      "loss": 6.504,
      "step": 63500
    },
    {
      "epoch": 0.87,
      "learning_rate": 8.48389134554643e-06,
      "loss": 6.5018,
      "step": 64000
    },
    {
      "epoch": 0.88,
      "learning_rate": 8.46985330244964e-06,
      "loss": 6.5007,
      "step": 64500
    },
    {
      "epoch": 0.89,
      "learning_rate": 8.455815259352846e-06,
      "loss": 6.4936,
      "step": 65000
    },
    {
      "epoch": 0.89,
      "learning_rate": 8.441777216256054e-06,
      "loss": 6.497,
      "step": 65500
    },
    {
      "epoch": 0.9,
      "learning_rate": 8.427739173159263e-06,
      "loss": 6.4917,
      "step": 66000
    },
    {
      "epoch": 0.91,
      "learning_rate": 8.413701130062471e-06,
      "loss": 6.491,
      "step": 66500
    },
    {
      "epoch": 0.91,
      "learning_rate": 8.399663086965678e-06,
      "loss": 6.4863,
      "step": 67000
    },
    {
      "epoch": 0.92,
      "learning_rate": 8.385625043868887e-06,
      "loss": 6.4871,
      "step": 67500
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.371587000772093e-06,
      "loss": 6.483,
      "step": 68000
    },
    {
      "epoch": 0.94,
      "learning_rate": 8.357548957675302e-06,
      "loss": 6.4847,
      "step": 68500
    },
    {
      "epoch": 0.94,
      "learning_rate": 8.343510914578508e-06,
      "loss": 6.4827,
      "step": 69000
    },
    {
      "epoch": 0.95,
      "learning_rate": 8.329472871481717e-06,
      "loss": 6.4851,
      "step": 69500
    },
    {
      "epoch": 0.96,
      "learning_rate": 8.315434828384924e-06,
      "loss": 6.479,
      "step": 70000
    },
    {
      "epoch": 0.96,
      "eval_accuracy": 0.12116699137234754,
      "eval_loss": 6.474371433258057,
      "eval_runtime": 592.8719,
      "eval_samples_per_second": 520.106,
      "eval_steps_per_second": 6.502,
      "step": 70000
    },
    {
      "epoch": 0.96,
      "learning_rate": 8.301396785288132e-06,
      "loss": 6.4718,
      "step": 70500
    },
    {
      "epoch": 0.97,
      "learning_rate": 8.287358742191339e-06,
      "loss": 6.4705,
      "step": 71000
    },
    {
      "epoch": 0.98,
      "learning_rate": 8.273320699094547e-06,
      "loss": 6.4778,
      "step": 71500
    },
    {
      "epoch": 0.98,
      "learning_rate": 8.259282655997754e-06,
      "loss": 6.4679,
      "step": 72000
    },
    {
      "epoch": 0.99,
      "learning_rate": 8.245244612900963e-06,
      "loss": 6.4698,
      "step": 72500
    },
    {
      "epoch": 1.0,
      "learning_rate": 8.23120656980417e-06,
      "loss": 6.464,
      "step": 73000
    },
    {
      "epoch": 1.0,
      "learning_rate": 8.217168526707378e-06,
      "loss": 6.4625,
      "step": 73500
    },
    {
      "epoch": 1.01,
      "learning_rate": 8.203130483610585e-06,
      "loss": 6.461,
      "step": 74000
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.189092440513793e-06,
      "loss": 6.4663,
      "step": 74500
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.175054397417e-06,
      "loss": 6.4606,
      "step": 75000
    },
    {
      "epoch": 1.03,
      "learning_rate": 8.161016354320208e-06,
      "loss": 6.4512,
      "step": 75500
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.146978311223415e-06,
      "loss": 6.4582,
      "step": 76000
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.132940268126623e-06,
      "loss": 6.4527,
      "step": 76500
    },
    {
      "epoch": 1.05,
      "learning_rate": 8.118902225029832e-06,
      "loss": 6.4498,
      "step": 77000
    },
    {
      "epoch": 1.06,
      "learning_rate": 8.104864181933039e-06,
      "loss": 6.4446,
      "step": 77500
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.090826138836247e-06,
      "loss": 6.4412,
      "step": 78000
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.076788095739456e-06,
      "loss": 6.4498,
      "step": 78500
    },
    {
      "epoch": 1.08,
      "learning_rate": 8.062750052642662e-06,
      "loss": 6.4467,
      "step": 79000
    },
    {
      "epoch": 1.09,
      "learning_rate": 8.04871200954587e-06,
      "loss": 6.4374,
      "step": 79500
    },
    {
      "epoch": 1.09,
      "learning_rate": 8.034673966449078e-06,
      "loss": 6.4372,
      "step": 80000
    },
    {
      "epoch": 1.09,
      "eval_accuracy": 0.12552207039740257,
      "eval_loss": 6.438162326812744,
      "eval_runtime": 594.671,
      "eval_samples_per_second": 518.532,
      "eval_steps_per_second": 6.483,
      "step": 80000
    },
    {
      "epoch": 1.1,
      "learning_rate": 8.020635923352286e-06,
      "loss": 6.4353,
      "step": 80500
    },
    {
      "epoch": 1.11,
      "learning_rate": 8.006597880255493e-06,
      "loss": 6.432,
      "step": 81000
    },
    {
      "epoch": 1.11,
      "learning_rate": 7.992559837158701e-06,
      "loss": 6.433,
      "step": 81500
    },
    {
      "epoch": 1.12,
      "learning_rate": 7.978521794061908e-06,
      "loss": 6.4293,
      "step": 82000
    },
    {
      "epoch": 1.13,
      "learning_rate": 7.964483750965116e-06,
      "loss": 6.4257,
      "step": 82500
    },
    {
      "epoch": 1.13,
      "learning_rate": 7.950445707868323e-06,
      "loss": 6.4276,
      "step": 83000
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.936407664771532e-06,
      "loss": 6.4269,
      "step": 83500
    },
    {
      "epoch": 1.15,
      "learning_rate": 7.92236962167474e-06,
      "loss": 6.424,
      "step": 84000
    },
    {
      "epoch": 1.15,
      "learning_rate": 7.908331578577947e-06,
      "loss": 6.4213,
      "step": 84500
    },
    {
      "epoch": 1.16,
      "learning_rate": 7.894293535481155e-06,
      "loss": 6.4219,
      "step": 85000
    },
    {
      "epoch": 1.17,
      "learning_rate": 7.880255492384362e-06,
      "loss": 6.4226,
      "step": 85500
    },
    {
      "epoch": 1.17,
      "learning_rate": 7.86621744928757e-06,
      "loss": 6.4223,
      "step": 86000
    },
    {
      "epoch": 1.18,
      "learning_rate": 7.852179406190777e-06,
      "loss": 6.4217,
      "step": 86500
    },
    {
      "epoch": 1.19,
      "learning_rate": 7.838141363093986e-06,
      "loss": 6.418,
      "step": 87000
    },
    {
      "epoch": 1.19,
      "learning_rate": 7.824103319997192e-06,
      "loss": 6.4234,
      "step": 87500
    },
    {
      "epoch": 1.2,
      "learning_rate": 7.810065276900401e-06,
      "loss": 6.4138,
      "step": 88000
    },
    {
      "epoch": 1.21,
      "learning_rate": 7.796027233803608e-06,
      "loss": 6.413,
      "step": 88500
    },
    {
      "epoch": 1.22,
      "learning_rate": 7.781989190706816e-06,
      "loss": 6.408,
      "step": 89000
    },
    {
      "epoch": 1.22,
      "learning_rate": 7.767951147610023e-06,
      "loss": 6.4122,
      "step": 89500
    },
    {
      "epoch": 1.23,
      "learning_rate": 7.753913104513231e-06,
      "loss": 6.4069,
      "step": 90000
    },
    {
      "epoch": 1.23,
      "eval_accuracy": 0.12752297866286239,
      "eval_loss": 6.40709924697876,
      "eval_runtime": 591.8179,
      "eval_samples_per_second": 521.032,
      "eval_steps_per_second": 6.514,
      "step": 90000
    },
    {
      "epoch": 1.24,
      "learning_rate": 7.739875061416438e-06,
      "loss": 6.4065,
      "step": 90500
    },
    {
      "epoch": 1.24,
      "learning_rate": 7.725837018319647e-06,
      "loss": 6.401,
      "step": 91000
    },
    {
      "epoch": 1.25,
      "learning_rate": 7.711798975222855e-06,
      "loss": 6.4051,
      "step": 91500
    },
    {
      "epoch": 1.26,
      "learning_rate": 7.697760932126063e-06,
      "loss": 6.4005,
      "step": 92000
    },
    {
      "epoch": 1.26,
      "learning_rate": 7.68372288902927e-06,
      "loss": 6.4007,
      "step": 92500
    },
    {
      "epoch": 1.27,
      "learning_rate": 7.669684845932479e-06,
      "loss": 6.3982,
      "step": 93000
    },
    {
      "epoch": 1.28,
      "learning_rate": 7.655646802835685e-06,
      "loss": 6.396,
      "step": 93500
    },
    {
      "epoch": 1.28,
      "learning_rate": 7.641608759738894e-06,
      "loss": 6.3954,
      "step": 94000
    },
    {
      "epoch": 1.29,
      "learning_rate": 7.627570716642101e-06,
      "loss": 6.3957,
      "step": 94500
    },
    {
      "epoch": 1.3,
      "learning_rate": 7.613532673545308e-06,
      "loss": 6.3896,
      "step": 95000
    },
    {
      "epoch": 1.3,
      "learning_rate": 7.599494630448516e-06,
      "loss": 6.396,
      "step": 95500
    },
    {
      "epoch": 1.31,
      "learning_rate": 7.5854565873517235e-06,
      "loss": 6.3903,
      "step": 96000
    },
    {
      "epoch": 1.32,
      "learning_rate": 7.571418544254932e-06,
      "loss": 6.3927,
      "step": 96500
    },
    {
      "epoch": 1.32,
      "learning_rate": 7.5573805011581395e-06,
      "loss": 6.3939,
      "step": 97000
    },
    {
      "epoch": 1.33,
      "learning_rate": 7.543342458061347e-06,
      "loss": 6.3883,
      "step": 97500
    },
    {
      "epoch": 1.34,
      "learning_rate": 7.529304414964555e-06,
      "loss": 6.3888,
      "step": 98000
    },
    {
      "epoch": 1.34,
      "learning_rate": 7.515266371867762e-06,
      "loss": 6.3884,
      "step": 98500
    },
    {
      "epoch": 1.35,
      "learning_rate": 7.50122832877097e-06,
      "loss": 6.3836,
      "step": 99000
    },
    {
      "epoch": 1.36,
      "learning_rate": 7.4871902856741775e-06,
      "loss": 6.3847,
      "step": 99500
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.473152242577385e-06,
      "loss": 6.3918,
      "step": 100000
    },
    {
      "epoch": 1.37,
      "eval_accuracy": 0.1302080024596662,
      "eval_loss": 6.380642414093018,
      "eval_runtime": 593.1372,
      "eval_samples_per_second": 519.873,
      "eval_steps_per_second": 6.499,
      "step": 100000
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.459114199480593e-06,
      "loss": 6.3838,
      "step": 100500
    },
    {
      "epoch": 1.38,
      "learning_rate": 7.4450761563838e-06,
      "loss": 6.382,
      "step": 101000
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.431038113287008e-06,
      "loss": 6.3773,
      "step": 101500
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.4170000701902156e-06,
      "loss": 6.3816,
      "step": 102000
    },
    {
      "epoch": 1.4,
      "learning_rate": 7.402962027093423e-06,
      "loss": 6.3813,
      "step": 102500
    },
    {
      "epoch": 1.41,
      "learning_rate": 7.388923983996632e-06,
      "loss": 6.3763,
      "step": 103000
    },
    {
      "epoch": 1.41,
      "learning_rate": 7.37488594089984e-06,
      "loss": 6.3775,
      "step": 103500
    },
    {
      "epoch": 1.42,
      "learning_rate": 7.360847897803048e-06,
      "loss": 6.377,
      "step": 104000
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.346809854706255e-06,
      "loss": 6.3717,
      "step": 104500
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.332771811609463e-06,
      "loss": 6.3746,
      "step": 105000
    },
    {
      "epoch": 1.44,
      "learning_rate": 7.3187337685126705e-06,
      "loss": 6.3664,
      "step": 105500
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.304695725415878e-06,
      "loss": 6.3681,
      "step": 106000
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.290657682319086e-06,
      "loss": 6.3747,
      "step": 106500
    },
    {
      "epoch": 1.46,
      "learning_rate": 7.276619639222293e-06,
      "loss": 6.363,
      "step": 107000
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.262581596125501e-06,
      "loss": 6.3671,
      "step": 107500
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.2485435530287085e-06,
      "loss": 6.3658,
      "step": 108000
    },
    {
      "epoch": 1.48,
      "learning_rate": 7.234505509931916e-06,
      "loss": 6.3624,
      "step": 108500
    },
    {
      "epoch": 1.49,
      "learning_rate": 7.220467466835124e-06,
      "loss": 6.3631,
      "step": 109000
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.206429423738331e-06,
      "loss": 6.363,
      "step": 109500
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.192391380641539e-06,
      "loss": 6.3625,
      "step": 110000
    },
    {
      "epoch": 1.5,
      "eval_accuracy": 0.1320408782502723,
      "eval_loss": 6.358105182647705,
      "eval_runtime": 594.9215,
      "eval_samples_per_second": 518.314,
      "eval_steps_per_second": 6.48,
      "step": 110000
    },
    {
      "epoch": 1.51,
      "learning_rate": 7.1783533375447466e-06,
      "loss": 6.357,
      "step": 110500
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.164315294447954e-06,
      "loss": 6.3529,
      "step": 111000
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.150277251351162e-06,
      "loss": 6.3628,
      "step": 111500
    },
    {
      "epoch": 1.53,
      "learning_rate": 7.136239208254369e-06,
      "loss": 6.3544,
      "step": 112000
    },
    {
      "epoch": 1.54,
      "learning_rate": 7.122201165157577e-06,
      "loss": 6.3506,
      "step": 112500
    },
    {
      "epoch": 1.54,
      "learning_rate": 7.108163122060785e-06,
      "loss": 6.3506,
      "step": 113000
    },
    {
      "epoch": 1.55,
      "learning_rate": 7.094125078963992e-06,
      "loss": 6.356,
      "step": 113500
    },
    {
      "epoch": 1.56,
      "learning_rate": 7.080087035867201e-06,
      "loss": 6.3543,
      "step": 114000
    },
    {
      "epoch": 1.56,
      "learning_rate": 7.066048992770408e-06,
      "loss": 6.345,
      "step": 114500
    },
    {
      "epoch": 1.57,
      "learning_rate": 7.052010949673616e-06,
      "loss": 6.3456,
      "step": 115000
    },
    {
      "epoch": 1.58,
      "learning_rate": 7.037972906576824e-06,
      "loss": 6.346,
      "step": 115500
    },
    {
      "epoch": 1.58,
      "learning_rate": 7.023934863480032e-06,
      "loss": 6.354,
      "step": 116000
    },
    {
      "epoch": 1.59,
      "learning_rate": 7.0098968203832395e-06,
      "loss": 6.3525,
      "step": 116500
    },
    {
      "epoch": 1.6,
      "learning_rate": 6.995858777286447e-06,
      "loss": 6.3504,
      "step": 117000
    },
    {
      "epoch": 1.6,
      "learning_rate": 6.981820734189655e-06,
      "loss": 6.3488,
      "step": 117500
    },
    {
      "epoch": 1.61,
      "learning_rate": 6.967782691092862e-06,
      "loss": 6.3461,
      "step": 118000
    },
    {
      "epoch": 1.62,
      "learning_rate": 6.95374464799607e-06,
      "loss": 6.3444,
      "step": 118500
    },
    {
      "epoch": 1.62,
      "learning_rate": 6.9397066048992776e-06,
      "loss": 6.3456,
      "step": 119000
    },
    {
      "epoch": 1.63,
      "learning_rate": 6.925668561802485e-06,
      "loss": 6.3397,
      "step": 119500
    },
    {
      "epoch": 1.64,
      "learning_rate": 6.911630518705693e-06,
      "loss": 6.339,
      "step": 120000
    },
    {
      "epoch": 1.64,
      "eval_accuracy": 0.13408538353915958,
      "eval_loss": 6.340002536773682,
      "eval_runtime": 593.3142,
      "eval_samples_per_second": 519.718,
      "eval_steps_per_second": 6.497,
      "step": 120000
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.8975924756089e-06,
      "loss": 6.3467,
      "step": 120500
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.883554432512109e-06,
      "loss": 6.3403,
      "step": 121000
    },
    {
      "epoch": 1.66,
      "learning_rate": 6.8695163894153164e-06,
      "loss": 6.342,
      "step": 121500
    },
    {
      "epoch": 1.67,
      "learning_rate": 6.855478346318524e-06,
      "loss": 6.3435,
      "step": 122000
    },
    {
      "epoch": 1.67,
      "learning_rate": 6.841440303221732e-06,
      "loss": 6.3354,
      "step": 122500
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.827402260124939e-06,
      "loss": 6.3382,
      "step": 123000
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.813364217028147e-06,
      "loss": 6.3381,
      "step": 123500
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.7993261739313545e-06,
      "loss": 6.3362,
      "step": 124000
    },
    {
      "epoch": 1.7,
      "learning_rate": 6.785288130834562e-06,
      "loss": 6.3347,
      "step": 124500
    },
    {
      "epoch": 1.71,
      "learning_rate": 6.77125008773777e-06,
      "loss": 6.3352,
      "step": 125000
    },
    {
      "epoch": 1.71,
      "learning_rate": 6.757212044640977e-06,
      "loss": 6.3303,
      "step": 125500
    },
    {
      "epoch": 1.72,
      "learning_rate": 6.743174001544185e-06,
      "loss": 6.3312,
      "step": 126000
    },
    {
      "epoch": 1.73,
      "learning_rate": 6.7291359584473925e-06,
      "loss": 6.3312,
      "step": 126500
    },
    {
      "epoch": 1.73,
      "learning_rate": 6.7150979153506e-06,
      "loss": 6.3258,
      "step": 127000
    },
    {
      "epoch": 1.74,
      "learning_rate": 6.701059872253808e-06,
      "loss": 6.3287,
      "step": 127500
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.687021829157016e-06,
      "loss": 6.3232,
      "step": 128000
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.672983786060225e-06,
      "loss": 6.3265,
      "step": 128500
    },
    {
      "epoch": 1.76,
      "learning_rate": 6.658945742963432e-06,
      "loss": 6.3246,
      "step": 129000
    },
    {
      "epoch": 1.77,
      "learning_rate": 6.64490769986664e-06,
      "loss": 6.3307,
      "step": 129500
    },
    {
      "epoch": 1.78,
      "learning_rate": 6.6308696567698474e-06,
      "loss": 6.3189,
      "step": 130000
    },
    {
      "epoch": 1.78,
      "eval_accuracy": 0.13503786641748597,
      "eval_loss": 6.3236799240112305,
      "eval_runtime": 593.509,
      "eval_samples_per_second": 519.547,
      "eval_steps_per_second": 6.495,
      "step": 130000
    },
    {
      "epoch": 1.78,
      "learning_rate": 6.616831613673055e-06,
      "loss": 6.3221,
      "step": 130500
    },
    {
      "epoch": 1.79,
      "learning_rate": 6.602793570576263e-06,
      "loss": 6.3188,
      "step": 131000
    },
    {
      "epoch": 1.8,
      "learning_rate": 6.58875552747947e-06,
      "loss": 6.3247,
      "step": 131500
    },
    {
      "epoch": 1.8,
      "learning_rate": 6.574717484382678e-06,
      "loss": 6.3225,
      "step": 132000
    },
    {
      "epoch": 1.81,
      "learning_rate": 6.5606794412858855e-06,
      "loss": 6.3199,
      "step": 132500
    },
    {
      "epoch": 1.82,
      "learning_rate": 6.546641398189093e-06,
      "loss": 6.3207,
      "step": 133000
    },
    {
      "epoch": 1.82,
      "learning_rate": 6.532603355092301e-06,
      "loss": 6.3189,
      "step": 133500
    },
    {
      "epoch": 1.83,
      "learning_rate": 6.518565311995508e-06,
      "loss": 6.318,
      "step": 134000
    },
    {
      "epoch": 1.84,
      "learning_rate": 6.504527268898716e-06,
      "loss": 6.3187,
      "step": 134500
    },
    {
      "epoch": 1.84,
      "learning_rate": 6.4904892258019235e-06,
      "loss": 6.3173,
      "step": 135000
    },
    {
      "epoch": 1.85,
      "learning_rate": 6.476451182705131e-06,
      "loss": 6.3184,
      "step": 135500
    },
    {
      "epoch": 1.86,
      "learning_rate": 6.462413139608339e-06,
      "loss": 6.3182,
      "step": 136000
    },
    {
      "epoch": 1.86,
      "learning_rate": 6.448375096511546e-06,
      "loss": 6.3178,
      "step": 136500
    },
    {
      "epoch": 1.87,
      "learning_rate": 6.434337053414754e-06,
      "loss": 6.3214,
      "step": 137000
    },
    {
      "epoch": 1.88,
      "learning_rate": 6.4202990103179615e-06,
      "loss": 6.3146,
      "step": 137500
    },
    {
      "epoch": 1.88,
      "learning_rate": 6.406260967221169e-06,
      "loss": 6.3159,
      "step": 138000
    },
    {
      "epoch": 1.89,
      "learning_rate": 6.392222924124378e-06,
      "loss": 6.3096,
      "step": 138500
    },
    {
      "epoch": 1.9,
      "learning_rate": 6.378184881027585e-06,
      "loss": 6.3126,
      "step": 139000
    },
    {
      "epoch": 1.9,
      "learning_rate": 6.364146837930793e-06,
      "loss": 6.3083,
      "step": 139500
    },
    {
      "epoch": 1.91,
      "learning_rate": 6.350108794834e-06,
      "loss": 6.3115,
      "step": 140000
    },
    {
      "epoch": 1.91,
      "eval_accuracy": 0.13663110510231702,
      "eval_loss": 6.310282230377197,
      "eval_runtime": 592.4399,
      "eval_samples_per_second": 520.485,
      "eval_steps_per_second": 6.507,
      "step": 140000
    },
    {
      "epoch": 1.92,
      "learning_rate": 6.336070751737209e-06,
      "loss": 6.3158,
      "step": 140500
    },
    {
      "epoch": 1.93,
      "learning_rate": 6.3220327086404165e-06,
      "loss": 6.3085,
      "step": 141000
    },
    {
      "epoch": 1.93,
      "learning_rate": 6.307994665543624e-06,
      "loss": 6.3147,
      "step": 141500
    },
    {
      "epoch": 1.94,
      "learning_rate": 6.293956622446832e-06,
      "loss": 6.31,
      "step": 142000
    },
    {
      "epoch": 1.95,
      "learning_rate": 6.279918579350039e-06,
      "loss": 6.3079,
      "step": 142500
    },
    {
      "epoch": 1.95,
      "learning_rate": 6.265880536253247e-06,
      "loss": 6.3082,
      "step": 143000
    },
    {
      "epoch": 1.96,
      "learning_rate": 6.2518424931564545e-06,
      "loss": 6.306,
      "step": 143500
    },
    {
      "epoch": 1.97,
      "learning_rate": 6.237804450059662e-06,
      "loss": 6.3103,
      "step": 144000
    },
    {
      "epoch": 1.97,
      "learning_rate": 6.22376640696287e-06,
      "loss": 6.3105,
      "step": 144500
    },
    {
      "epoch": 1.98,
      "learning_rate": 6.209728363866077e-06,
      "loss": 6.3033,
      "step": 145000
    },
    {
      "epoch": 1.99,
      "learning_rate": 6.195690320769285e-06,
      "loss": 6.3051,
      "step": 145500
    },
    {
      "epoch": 1.99,
      "learning_rate": 6.181652277672493e-06,
      "loss": 6.3027,
      "step": 146000
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.167614234575701e-06,
      "loss": 6.3016,
      "step": 146500
    },
    {
      "epoch": 2.01,
      "learning_rate": 6.153576191478909e-06,
      "loss": 6.2989,
      "step": 147000
    },
    {
      "epoch": 2.01,
      "learning_rate": 6.139538148382116e-06,
      "loss": 6.3049,
      "step": 147500
    },
    {
      "epoch": 2.02,
      "learning_rate": 6.125500105285324e-06,
      "loss": 6.2975,
      "step": 148000
    },
    {
      "epoch": 2.03,
      "learning_rate": 6.111462062188531e-06,
      "loss": 6.304,
      "step": 148500
    },
    {
      "epoch": 2.03,
      "learning_rate": 6.097424019091739e-06,
      "loss": 6.2981,
      "step": 149000
    },
    {
      "epoch": 2.04,
      "learning_rate": 6.083385975994947e-06,
      "loss": 6.2915,
      "step": 149500
    },
    {
      "epoch": 2.05,
      "learning_rate": 6.069347932898154e-06,
      "loss": 6.298,
      "step": 150000
    },
    {
      "epoch": 2.05,
      "eval_accuracy": 0.1374457755307256,
      "eval_loss": 6.298206806182861,
      "eval_runtime": 595.7013,
      "eval_samples_per_second": 517.635,
      "eval_steps_per_second": 6.471,
      "step": 150000
    },
    {
      "epoch": 2.06,
      "learning_rate": 6.055309889801362e-06,
      "loss": 6.2936,
      "step": 150500
    },
    {
      "epoch": 2.06,
      "learning_rate": 6.0412718467045694e-06,
      "loss": 6.3001,
      "step": 151000
    },
    {
      "epoch": 2.07,
      "learning_rate": 6.027233803607777e-06,
      "loss": 6.2945,
      "step": 151500
    },
    {
      "epoch": 2.08,
      "learning_rate": 6.013195760510985e-06,
      "loss": 6.2997,
      "step": 152000
    },
    {
      "epoch": 2.08,
      "learning_rate": 5.999157717414193e-06,
      "loss": 6.305,
      "step": 152500
    },
    {
      "epoch": 2.09,
      "learning_rate": 5.9851196743174016e-06,
      "loss": 6.2981,
      "step": 153000
    },
    {
      "epoch": 2.1,
      "learning_rate": 5.971081631220609e-06,
      "loss": 6.2951,
      "step": 153500
    },
    {
      "epoch": 2.1,
      "learning_rate": 5.957043588123817e-06,
      "loss": 6.2988,
      "step": 154000
    },
    {
      "epoch": 2.11,
      "learning_rate": 5.943005545027024e-06,
      "loss": 6.2923,
      "step": 154500
    },
    {
      "epoch": 2.12,
      "learning_rate": 5.928967501930232e-06,
      "loss": 6.2942,
      "step": 155000
    },
    {
      "epoch": 2.12,
      "learning_rate": 5.91492945883344e-06,
      "loss": 6.2914,
      "step": 155500
    },
    {
      "epoch": 2.13,
      "learning_rate": 5.900891415736647e-06,
      "loss": 6.2946,
      "step": 156000
    },
    {
      "epoch": 2.14,
      "learning_rate": 5.886853372639855e-06,
      "loss": 6.2947,
      "step": 156500
    },
    {
      "epoch": 2.14,
      "learning_rate": 5.872815329543062e-06,
      "loss": 6.2958,
      "step": 157000
    },
    {
      "epoch": 2.15,
      "learning_rate": 5.85877728644627e-06,
      "loss": 6.287,
      "step": 157500
    },
    {
      "epoch": 2.16,
      "learning_rate": 5.844739243349478e-06,
      "loss": 6.2903,
      "step": 158000
    },
    {
      "epoch": 2.16,
      "learning_rate": 5.830701200252685e-06,
      "loss": 6.2876,
      "step": 158500
    },
    {
      "epoch": 2.17,
      "learning_rate": 5.816663157155893e-06,
      "loss": 6.2881,
      "step": 159000
    },
    {
      "epoch": 2.18,
      "learning_rate": 5.8026251140591004e-06,
      "loss": 6.2902,
      "step": 159500
    },
    {
      "epoch": 2.18,
      "learning_rate": 5.788587070962308e-06,
      "loss": 6.2888,
      "step": 160000
    },
    {
      "epoch": 2.18,
      "eval_accuracy": 0.13820078499319666,
      "eval_loss": 6.287766456604004,
      "eval_runtime": 594.2657,
      "eval_samples_per_second": 518.886,
      "eval_steps_per_second": 6.487,
      "step": 160000
    },
    {
      "epoch": 2.19,
      "learning_rate": 5.774549027865516e-06,
      "loss": 6.2871,
      "step": 160500
    },
    {
      "epoch": 2.2,
      "learning_rate": 5.760510984768723e-06,
      "loss": 6.2971,
      "step": 161000
    },
    {
      "epoch": 2.21,
      "learning_rate": 5.746472941671931e-06,
      "loss": 6.2895,
      "step": 161500
    },
    {
      "epoch": 2.21,
      "learning_rate": 5.7324348985751385e-06,
      "loss": 6.2863,
      "step": 162000
    },
    {
      "epoch": 2.22,
      "learning_rate": 5.718396855478346e-06,
      "loss": 6.2878,
      "step": 162500
    },
    {
      "epoch": 2.23,
      "learning_rate": 5.704358812381554e-06,
      "loss": 6.2825,
      "step": 163000
    },
    {
      "epoch": 2.23,
      "learning_rate": 5.690320769284762e-06,
      "loss": 6.2872,
      "step": 163500
    },
    {
      "epoch": 2.24,
      "learning_rate": 5.67628272618797e-06,
      "loss": 6.2868,
      "step": 164000
    },
    {
      "epoch": 2.25,
      "learning_rate": 5.662244683091177e-06,
      "loss": 6.2763,
      "step": 164500
    },
    {
      "epoch": 2.25,
      "learning_rate": 5.648206639994386e-06,
      "loss": 6.2837,
      "step": 165000
    },
    {
      "epoch": 2.26,
      "learning_rate": 5.634168596897593e-06,
      "loss": 6.2821,
      "step": 165500
    },
    {
      "epoch": 2.27,
      "learning_rate": 5.620130553800801e-06,
      "loss": 6.2809,
      "step": 166000
    },
    {
      "epoch": 2.27,
      "learning_rate": 5.606092510704009e-06,
      "loss": 6.2852,
      "step": 166500
    },
    {
      "epoch": 2.28,
      "learning_rate": 5.592054467607216e-06,
      "loss": 6.2806,
      "step": 167000
    },
    {
      "epoch": 2.29,
      "learning_rate": 5.578016424510424e-06,
      "loss": 6.2828,
      "step": 167500
    },
    {
      "epoch": 2.29,
      "learning_rate": 5.5639783814136314e-06,
      "loss": 6.2778,
      "step": 168000
    },
    {
      "epoch": 2.3,
      "learning_rate": 5.549940338316839e-06,
      "loss": 6.2781,
      "step": 168500
    },
    {
      "epoch": 2.31,
      "learning_rate": 5.535902295220047e-06,
      "loss": 6.2773,
      "step": 169000
    },
    {
      "epoch": 2.31,
      "learning_rate": 5.521864252123254e-06,
      "loss": 6.2806,
      "step": 169500
    },
    {
      "epoch": 2.32,
      "learning_rate": 5.507826209026462e-06,
      "loss": 6.2786,
      "step": 170000
    },
    {
      "epoch": 2.32,
      "eval_accuracy": 0.1391640348507614,
      "eval_loss": 6.27783727645874,
      "eval_runtime": 591.2931,
      "eval_samples_per_second": 521.494,
      "eval_steps_per_second": 6.52,
      "step": 170000
    },
    {
      "epoch": 2.33,
      "learning_rate": 5.49378816592967e-06,
      "loss": 6.2786,
      "step": 170500
    },
    {
      "epoch": 2.33,
      "learning_rate": 5.479750122832878e-06,
      "loss": 6.2774,
      "step": 171000
    },
    {
      "epoch": 2.34,
      "learning_rate": 5.4657120797360855e-06,
      "loss": 6.2795,
      "step": 171500
    },
    {
      "epoch": 2.35,
      "learning_rate": 5.451674036639293e-06,
      "loss": 6.2817,
      "step": 172000
    },
    {
      "epoch": 2.36,
      "learning_rate": 5.437635993542501e-06,
      "loss": 6.2769,
      "step": 172500
    },
    {
      "epoch": 2.36,
      "learning_rate": 5.423597950445708e-06,
      "loss": 6.2726,
      "step": 173000
    },
    {
      "epoch": 2.37,
      "learning_rate": 5.409559907348916e-06,
      "loss": 6.2802,
      "step": 173500
    },
    {
      "epoch": 2.38,
      "learning_rate": 5.3955218642521235e-06,
      "loss": 6.2742,
      "step": 174000
    },
    {
      "epoch": 2.38,
      "learning_rate": 5.381483821155331e-06,
      "loss": 6.2707,
      "step": 174500
    },
    {
      "epoch": 2.39,
      "learning_rate": 5.367445778058539e-06,
      "loss": 6.2725,
      "step": 175000
    },
    {
      "epoch": 2.4,
      "learning_rate": 5.353407734961746e-06,
      "loss": 6.2715,
      "step": 175500
    },
    {
      "epoch": 2.4,
      "learning_rate": 5.339369691864954e-06,
      "loss": 6.2714,
      "step": 176000
    },
    {
      "epoch": 2.41,
      "learning_rate": 5.325331648768162e-06,
      "loss": 6.2759,
      "step": 176500
    },
    {
      "epoch": 2.42,
      "learning_rate": 5.311293605671369e-06,
      "loss": 6.2699,
      "step": 177000
    },
    {
      "epoch": 2.42,
      "learning_rate": 5.2972555625745785e-06,
      "loss": 6.2708,
      "step": 177500
    },
    {
      "epoch": 2.43,
      "learning_rate": 5.283217519477786e-06,
      "loss": 6.2705,
      "step": 178000
    },
    {
      "epoch": 2.44,
      "learning_rate": 5.269179476380994e-06,
      "loss": 6.2779,
      "step": 178500
    },
    {
      "epoch": 2.44,
      "learning_rate": 5.255141433284201e-06,
      "loss": 6.2704,
      "step": 179000
    },
    {
      "epoch": 2.45,
      "learning_rate": 5.241103390187409e-06,
      "loss": 6.2725,
      "step": 179500
    },
    {
      "epoch": 2.46,
      "learning_rate": 5.2270653470906165e-06,
      "loss": 6.2691,
      "step": 180000
    },
    {
      "epoch": 2.46,
      "eval_accuracy": 0.1397743053749513,
      "eval_loss": 6.270318984985352,
      "eval_runtime": 592.4485,
      "eval_samples_per_second": 520.477,
      "eval_steps_per_second": 6.507,
      "step": 180000
    },
    {
      "epoch": 2.46,
      "learning_rate": 5.213027303993824e-06,
      "loss": 6.2701,
      "step": 180500
    },
    {
      "epoch": 2.47,
      "learning_rate": 5.198989260897032e-06,
      "loss": 6.2748,
      "step": 181000
    },
    {
      "epoch": 2.48,
      "learning_rate": 5.184951217800239e-06,
      "loss": 6.2678,
      "step": 181500
    },
    {
      "epoch": 2.49,
      "learning_rate": 5.170913174703447e-06,
      "loss": 6.2689,
      "step": 182000
    },
    {
      "epoch": 2.49,
      "learning_rate": 5.1568751316066545e-06,
      "loss": 6.2711,
      "step": 182500
    },
    {
      "epoch": 2.5,
      "learning_rate": 5.142837088509862e-06,
      "loss": 6.2683,
      "step": 183000
    },
    {
      "epoch": 2.51,
      "learning_rate": 5.12879904541307e-06,
      "loss": 6.262,
      "step": 183500
    },
    {
      "epoch": 2.51,
      "learning_rate": 5.114761002316277e-06,
      "loss": 6.2625,
      "step": 184000
    },
    {
      "epoch": 2.52,
      "learning_rate": 5.100722959219485e-06,
      "loss": 6.2626,
      "step": 184500
    },
    {
      "epoch": 2.53,
      "learning_rate": 5.0866849161226926e-06,
      "loss": 6.2657,
      "step": 185000
    },
    {
      "epoch": 2.53,
      "learning_rate": 5.0726468730259e-06,
      "loss": 6.2547,
      "step": 185500
    },
    {
      "epoch": 2.54,
      "learning_rate": 5.058608829929108e-06,
      "loss": 6.266,
      "step": 186000
    },
    {
      "epoch": 2.55,
      "learning_rate": 5.044570786832315e-06,
      "loss": 6.2671,
      "step": 186500
    },
    {
      "epoch": 2.55,
      "learning_rate": 5.030532743735523e-06,
      "loss": 6.2653,
      "step": 187000
    },
    {
      "epoch": 2.56,
      "learning_rate": 5.016494700638731e-06,
      "loss": 6.266,
      "step": 187500
    },
    {
      "epoch": 2.57,
      "learning_rate": 5.002456657541939e-06,
      "loss": 6.2671,
      "step": 188000
    },
    {
      "epoch": 2.57,
      "learning_rate": 4.988418614445147e-06,
      "loss": 6.2659,
      "step": 188500
    },
    {
      "epoch": 2.58,
      "learning_rate": 4.974380571348354e-06,
      "loss": 6.2684,
      "step": 189000
    },
    {
      "epoch": 2.59,
      "learning_rate": 4.960342528251562e-06,
      "loss": 6.264,
      "step": 189500
    },
    {
      "epoch": 2.59,
      "learning_rate": 4.9463044851547695e-06,
      "loss": 6.2675,
      "step": 190000
    },
    {
      "epoch": 2.59,
      "eval_accuracy": 0.14022267505641858,
      "eval_loss": 6.2622222900390625,
      "eval_runtime": 592.3438,
      "eval_samples_per_second": 520.569,
      "eval_steps_per_second": 6.508,
      "step": 190000
    },
    {
      "epoch": 2.6,
      "learning_rate": 4.932266442057977e-06,
      "loss": 6.2617,
      "step": 190500
    },
    {
      "epoch": 2.61,
      "learning_rate": 4.918228398961185e-06,
      "loss": 6.2544,
      "step": 191000
    },
    {
      "epoch": 2.61,
      "learning_rate": 4.904190355864393e-06,
      "loss": 6.256,
      "step": 191500
    },
    {
      "epoch": 2.62,
      "learning_rate": 4.890152312767601e-06,
      "loss": 6.2589,
      "step": 192000
    },
    {
      "epoch": 2.63,
      "learning_rate": 4.876114269670808e-06,
      "loss": 6.2611,
      "step": 192500
    },
    {
      "epoch": 2.64,
      "learning_rate": 4.862076226574016e-06,
      "loss": 6.2618,
      "step": 193000
    },
    {
      "epoch": 2.64,
      "learning_rate": 4.8480381834772236e-06,
      "loss": 6.263,
      "step": 193500
    },
    {
      "epoch": 2.65,
      "learning_rate": 4.834000140380431e-06,
      "loss": 6.264,
      "step": 194000
    },
    {
      "epoch": 2.66,
      "learning_rate": 4.819962097283639e-06,
      "loss": 6.259,
      "step": 194500
    },
    {
      "epoch": 2.66,
      "learning_rate": 4.805924054186847e-06,
      "loss": 6.2627,
      "step": 195000
    },
    {
      "epoch": 2.67,
      "learning_rate": 4.791886011090055e-06,
      "loss": 6.2586,
      "step": 195500
    },
    {
      "epoch": 2.68,
      "learning_rate": 4.7778479679932624e-06,
      "loss": 6.2588,
      "step": 196000
    },
    {
      "epoch": 2.68,
      "learning_rate": 4.76380992489647e-06,
      "loss": 6.2584,
      "step": 196500
    },
    {
      "epoch": 2.69,
      "learning_rate": 4.749771881799678e-06,
      "loss": 6.2585,
      "step": 197000
    },
    {
      "epoch": 2.7,
      "learning_rate": 4.735733838702885e-06,
      "loss": 6.2575,
      "step": 197500
    },
    {
      "epoch": 2.7,
      "learning_rate": 4.721695795606093e-06,
      "loss": 6.2514,
      "step": 198000
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.7076577525093005e-06,
      "loss": 6.2519,
      "step": 198500
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.693619709412509e-06,
      "loss": 6.2572,
      "step": 199000
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.6795816663157165e-06,
      "loss": 6.2567,
      "step": 199500
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.665543623218924e-06,
      "loss": 6.255,
      "step": 200000
    },
    {
      "epoch": 2.73,
      "eval_accuracy": 0.14042373264472863,
      "eval_loss": 6.257112503051758,
      "eval_runtime": 591.5678,
      "eval_samples_per_second": 521.252,
      "eval_steps_per_second": 6.517,
      "step": 200000
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.651505580122132e-06,
      "loss": 6.2588,
      "step": 200500
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.637467537025339e-06,
      "loss": 6.2547,
      "step": 201000
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.623429493928547e-06,
      "loss": 6.2597,
      "step": 201500
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.6093914508317546e-06,
      "loss": 6.2566,
      "step": 202000
    },
    {
      "epoch": 2.77,
      "learning_rate": 4.595353407734962e-06,
      "loss": 6.255,
      "step": 202500
    },
    {
      "epoch": 2.77,
      "learning_rate": 4.58131536463817e-06,
      "loss": 6.2521,
      "step": 203000
    },
    {
      "epoch": 2.78,
      "learning_rate": 4.567277321541377e-06,
      "loss": 6.2588,
      "step": 203500
    },
    {
      "epoch": 2.79,
      "learning_rate": 4.553239278444585e-06,
      "loss": 6.2518,
      "step": 204000
    },
    {
      "epoch": 2.79,
      "learning_rate": 4.539201235347793e-06,
      "loss": 6.252,
      "step": 204500
    },
    {
      "epoch": 2.8,
      "learning_rate": 4.525163192251e-06,
      "loss": 6.2515,
      "step": 205000
    },
    {
      "epoch": 2.81,
      "learning_rate": 4.511125149154209e-06,
      "loss": 6.2537,
      "step": 205500
    },
    {
      "epoch": 2.81,
      "learning_rate": 4.497087106057416e-06,
      "loss": 6.2566,
      "step": 206000
    },
    {
      "epoch": 2.82,
      "learning_rate": 4.483049062960624e-06,
      "loss": 6.2497,
      "step": 206500
    },
    {
      "epoch": 2.83,
      "learning_rate": 4.4690110198638315e-06,
      "loss": 6.2539,
      "step": 207000
    },
    {
      "epoch": 2.83,
      "learning_rate": 4.454972976767039e-06,
      "loss": 6.2576,
      "step": 207500
    },
    {
      "epoch": 2.84,
      "learning_rate": 4.440934933670247e-06,
      "loss": 6.2476,
      "step": 208000
    },
    {
      "epoch": 2.85,
      "learning_rate": 4.426896890573454e-06,
      "loss": 6.2469,
      "step": 208500
    },
    {
      "epoch": 2.85,
      "learning_rate": 4.412858847476662e-06,
      "loss": 6.2482,
      "step": 209000
    },
    {
      "epoch": 2.86,
      "learning_rate": 4.3988208043798695e-06,
      "loss": 6.2538,
      "step": 209500
    },
    {
      "epoch": 2.87,
      "learning_rate": 4.384782761283077e-06,
      "loss": 6.2481,
      "step": 210000
    },
    {
      "epoch": 2.87,
      "eval_accuracy": 0.14072426813488315,
      "eval_loss": 6.251694202423096,
      "eval_runtime": 593.3315,
      "eval_samples_per_second": 519.703,
      "eval_steps_per_second": 6.497,
      "step": 210000
    },
    {
      "epoch": 2.87,
      "learning_rate": 4.370744718186285e-06,
      "loss": 6.2527,
      "step": 210500
    },
    {
      "epoch": 2.88,
      "learning_rate": 4.356706675089492e-06,
      "loss": 6.2581,
      "step": 211000
    },
    {
      "epoch": 2.89,
      "learning_rate": 4.342668631992701e-06,
      "loss": 6.2492,
      "step": 211500
    },
    {
      "epoch": 2.89,
      "learning_rate": 4.328630588895908e-06,
      "loss": 6.2542,
      "step": 212000
    },
    {
      "epoch": 2.9,
      "learning_rate": 4.314592545799116e-06,
      "loss": 6.2525,
      "step": 212500
    },
    {
      "epoch": 2.91,
      "learning_rate": 4.300554502702324e-06,
      "loss": 6.2528,
      "step": 213000
    },
    {
      "epoch": 2.92,
      "learning_rate": 4.286516459605531e-06,
      "loss": 6.2521,
      "step": 213500
    },
    {
      "epoch": 2.92,
      "learning_rate": 4.272478416508739e-06,
      "loss": 6.2533,
      "step": 214000
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.258440373411946e-06,
      "loss": 6.2483,
      "step": 214500
    },
    {
      "epoch": 2.94,
      "learning_rate": 4.244402330315154e-06,
      "loss": 6.2484,
      "step": 215000
    },
    {
      "epoch": 2.94,
      "learning_rate": 4.230364287218362e-06,
      "loss": 6.248,
      "step": 215500
    },
    {
      "epoch": 2.95,
      "learning_rate": 4.21632624412157e-06,
      "loss": 6.2464,
      "step": 216000
    },
    {
      "epoch": 2.96,
      "learning_rate": 4.202288201024778e-06,
      "loss": 6.2531,
      "step": 216500
    },
    {
      "epoch": 2.96,
      "learning_rate": 4.188250157927985e-06,
      "loss": 6.2456,
      "step": 217000
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.174212114831193e-06,
      "loss": 6.2425,
      "step": 217500
    },
    {
      "epoch": 2.98,
      "learning_rate": 4.1601740717344005e-06,
      "loss": 6.2481,
      "step": 218000
    },
    {
      "epoch": 2.98,
      "learning_rate": 4.146136028637608e-06,
      "loss": 6.2478,
      "step": 218500
    },
    {
      "epoch": 2.99,
      "learning_rate": 4.132097985540816e-06,
      "loss": 6.2478,
      "step": 219000
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.118059942444023e-06,
      "loss": 6.2478,
      "step": 219500
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.104021899347232e-06,
      "loss": 6.2465,
      "step": 220000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.1416870933675662,
      "eval_loss": 6.244670867919922,
      "eval_runtime": 595.9761,
      "eval_samples_per_second": 517.397,
      "eval_steps_per_second": 6.468,
      "step": 220000
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.089983856250439e-06,
      "loss": 6.2446,
      "step": 220500
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.075945813153647e-06,
      "loss": 6.2479,
      "step": 221000
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.061907770056855e-06,
      "loss": 6.2429,
      "step": 221500
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.047869726960062e-06,
      "loss": 6.2495,
      "step": 222000
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.03383168386327e-06,
      "loss": 6.2461,
      "step": 222500
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.019793640766477e-06,
      "loss": 6.248,
      "step": 223000
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.005755597669685e-06,
      "loss": 6.2443,
      "step": 223500
    },
    {
      "epoch": 3.06,
      "learning_rate": 3.9917175545728935e-06,
      "loss": 6.2455,
      "step": 224000
    },
    {
      "epoch": 3.07,
      "learning_rate": 3.977679511476101e-06,
      "loss": 6.2372,
      "step": 224500
    },
    {
      "epoch": 3.07,
      "learning_rate": 3.963641468379309e-06,
      "loss": 6.2418,
      "step": 225000
    },
    {
      "epoch": 3.08,
      "learning_rate": 3.949603425282516e-06,
      "loss": 6.2408,
      "step": 225500
    },
    {
      "epoch": 3.09,
      "learning_rate": 3.935565382185724e-06,
      "loss": 6.2438,
      "step": 226000
    },
    {
      "epoch": 3.09,
      "learning_rate": 3.9215273390889315e-06,
      "loss": 6.242,
      "step": 226500
    },
    {
      "epoch": 3.1,
      "learning_rate": 3.907489295992139e-06,
      "loss": 6.2454,
      "step": 227000
    },
    {
      "epoch": 3.11,
      "learning_rate": 3.893451252895347e-06,
      "loss": 6.2407,
      "step": 227500
    },
    {
      "epoch": 3.11,
      "learning_rate": 3.879413209798554e-06,
      "loss": 6.2391,
      "step": 228000
    },
    {
      "epoch": 3.12,
      "learning_rate": 3.865375166701762e-06,
      "loss": 6.2389,
      "step": 228500
    },
    {
      "epoch": 3.13,
      "learning_rate": 3.8513371236049695e-06,
      "loss": 6.2429,
      "step": 229000
    },
    {
      "epoch": 3.13,
      "learning_rate": 3.837299080508177e-06,
      "loss": 6.2386,
      "step": 229500
    },
    {
      "epoch": 3.14,
      "learning_rate": 3.823261037411386e-06,
      "loss": 6.2381,
      "step": 230000
    },
    {
      "epoch": 3.14,
      "eval_accuracy": 0.14192931802585865,
      "eval_loss": 6.239927768707275,
      "eval_runtime": 593.8509,
      "eval_samples_per_second": 519.248,
      "eval_steps_per_second": 6.492,
      "step": 230000
    },
    {
      "epoch": 3.15,
      "learning_rate": 3.809222994314593e-06,
      "loss": 6.2364,
      "step": 230500
    },
    {
      "epoch": 3.15,
      "learning_rate": 3.795184951217801e-06,
      "loss": 6.2454,
      "step": 231000
    },
    {
      "epoch": 3.16,
      "learning_rate": 3.7811469081210084e-06,
      "loss": 6.2317,
      "step": 231500
    },
    {
      "epoch": 3.17,
      "learning_rate": 3.767108865024216e-06,
      "loss": 6.237,
      "step": 232000
    },
    {
      "epoch": 3.17,
      "learning_rate": 3.7530708219274236e-06,
      "loss": 6.2369,
      "step": 232500
    },
    {
      "epoch": 3.18,
      "learning_rate": 3.7390327788306312e-06,
      "loss": 6.2348,
      "step": 233000
    },
    {
      "epoch": 3.19,
      "learning_rate": 3.724994735733839e-06,
      "loss": 6.2427,
      "step": 233500
    },
    {
      "epoch": 3.2,
      "learning_rate": 3.7109566926370464e-06,
      "loss": 6.2359,
      "step": 234000
    },
    {
      "epoch": 3.2,
      "learning_rate": 3.6969186495402545e-06,
      "loss": 6.2417,
      "step": 234500
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.682880606443462e-06,
      "loss": 6.2406,
      "step": 235000
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.6688425633466697e-06,
      "loss": 6.2387,
      "step": 235500
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.6548045202498773e-06,
      "loss": 6.236,
      "step": 236000
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.6407664771530853e-06,
      "loss": 6.2367,
      "step": 236500
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.626728434056293e-06,
      "loss": 6.2375,
      "step": 237000
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.6126903909595005e-06,
      "loss": 6.2404,
      "step": 237500
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.598652347862708e-06,
      "loss": 6.2354,
      "step": 238000
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.584614304765916e-06,
      "loss": 6.2338,
      "step": 238500
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.5705762616691238e-06,
      "loss": 6.2376,
      "step": 239000
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.5565382185723314e-06,
      "loss": 6.2394,
      "step": 239500
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.542500175475539e-06,
      "loss": 6.2332,
      "step": 240000
    },
    {
      "epoch": 3.28,
      "eval_accuracy": 0.14218110691264726,
      "eval_loss": 6.237051010131836,
      "eval_runtime": 594.715,
      "eval_samples_per_second": 518.494,
      "eval_steps_per_second": 6.482,
      "step": 240000
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.5284621323787466e-06,
      "loss": 6.2367,
      "step": 240500
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.514424089281954e-06,
      "loss": 6.2401,
      "step": 241000
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.500386046185162e-06,
      "loss": 6.2311,
      "step": 241500
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.4863480030883694e-06,
      "loss": 6.2429,
      "step": 242000
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.472309959991578e-06,
      "loss": 6.2413,
      "step": 242500
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.4582719168947855e-06,
      "loss": 6.2376,
      "step": 243000
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.444233873797993e-06,
      "loss": 6.2352,
      "step": 243500
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.4301958307012007e-06,
      "loss": 6.2324,
      "step": 244000
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.4161577876044083e-06,
      "loss": 6.2335,
      "step": 244500
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.402119744507616e-06,
      "loss": 6.2349,
      "step": 245000
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.3880817014108235e-06,
      "loss": 6.2387,
      "step": 245500
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.374043658314031e-06,
      "loss": 6.2311,
      "step": 246000
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.3600056152172387e-06,
      "loss": 6.2319,
      "step": 246500
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.3459675721204467e-06,
      "loss": 6.2363,
      "step": 247000
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.3319295290236543e-06,
      "loss": 6.2307,
      "step": 247500
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.317891485926862e-06,
      "loss": 6.2336,
      "step": 248000
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.30385344283007e-06,
      "loss": 6.2291,
      "step": 248500
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.2898153997332776e-06,
      "loss": 6.2338,
      "step": 249000
    },
    {
      "epoch": 3.41,
      "learning_rate": 3.275777356636485e-06,
      "loss": 6.237,
      "step": 249500
    },
    {
      "epoch": 3.41,
      "learning_rate": 3.261739313539693e-06,
      "loss": 6.2279,
      "step": 250000
    },
    {
      "epoch": 3.41,
      "eval_accuracy": 0.14244212885752025,
      "eval_loss": 6.23211669921875,
      "eval_runtime": 593.6522,
      "eval_samples_per_second": 519.422,
      "eval_steps_per_second": 6.494,
      "step": 250000
    },
    {
      "epoch": 3.42,
      "learning_rate": 3.247701270442901e-06,
      "loss": 6.2298,
      "step": 250500
    },
    {
      "epoch": 3.43,
      "learning_rate": 3.2336632273461084e-06,
      "loss": 6.2368,
      "step": 251000
    },
    {
      "epoch": 3.43,
      "learning_rate": 3.219625184249316e-06,
      "loss": 6.2307,
      "step": 251500
    },
    {
      "epoch": 3.44,
      "learning_rate": 3.2055871411525236e-06,
      "loss": 6.2318,
      "step": 252000
    },
    {
      "epoch": 3.45,
      "learning_rate": 3.1915490980557312e-06,
      "loss": 6.2284,
      "step": 252500
    },
    {
      "epoch": 3.45,
      "learning_rate": 3.177511054958939e-06,
      "loss": 6.2331,
      "step": 253000
    },
    {
      "epoch": 3.46,
      "learning_rate": 3.1634730118621465e-06,
      "loss": 6.2283,
      "step": 253500
    },
    {
      "epoch": 3.47,
      "learning_rate": 3.149434968765354e-06,
      "loss": 6.2319,
      "step": 254000
    },
    {
      "epoch": 3.48,
      "learning_rate": 3.1353969256685617e-06,
      "loss": 6.237,
      "step": 254500
    },
    {
      "epoch": 3.48,
      "learning_rate": 3.12135888257177e-06,
      "loss": 6.2353,
      "step": 255000
    },
    {
      "epoch": 3.49,
      "learning_rate": 3.1073208394749777e-06,
      "loss": 6.2289,
      "step": 255500
    },
    {
      "epoch": 3.5,
      "learning_rate": 3.0932827963781853e-06,
      "loss": 6.2304,
      "step": 256000
    },
    {
      "epoch": 3.5,
      "learning_rate": 3.079244753281393e-06,
      "loss": 6.227,
      "step": 256500
    },
    {
      "epoch": 3.51,
      "learning_rate": 3.0652067101846005e-06,
      "loss": 6.2303,
      "step": 257000
    },
    {
      "epoch": 3.52,
      "learning_rate": 3.051168667087808e-06,
      "loss": 6.2283,
      "step": 257500
    },
    {
      "epoch": 3.52,
      "learning_rate": 3.0371306239910158e-06,
      "loss": 6.2286,
      "step": 258000
    },
    {
      "epoch": 3.53,
      "learning_rate": 3.0230925808942234e-06,
      "loss": 6.2377,
      "step": 258500
    },
    {
      "epoch": 3.54,
      "learning_rate": 3.009054537797431e-06,
      "loss": 6.232,
      "step": 259000
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.995016494700639e-06,
      "loss": 6.2296,
      "step": 259500
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.9809784516038466e-06,
      "loss": 6.2275,
      "step": 260000
    },
    {
      "epoch": 3.55,
      "eval_accuracy": 0.1426943792078911,
      "eval_loss": 6.227406978607178,
      "eval_runtime": 593.1661,
      "eval_samples_per_second": 519.848,
      "eval_steps_per_second": 6.499,
      "step": 260000
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.9669404085070542e-06,
      "loss": 6.2324,
      "step": 260500
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.9529023654102622e-06,
      "loss": 6.2316,
      "step": 261000
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.93886432231347e-06,
      "loss": 6.2233,
      "step": 261500
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.9248262792166775e-06,
      "loss": 6.2265,
      "step": 262000
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.910788236119885e-06,
      "loss": 6.2306,
      "step": 262500
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.896750193023093e-06,
      "loss": 6.2241,
      "step": 263000
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.8827121499263007e-06,
      "loss": 6.2323,
      "step": 263500
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.8686741068295083e-06,
      "loss": 6.229,
      "step": 264000
    },
    {
      "epoch": 3.61,
      "learning_rate": 2.854636063732716e-06,
      "loss": 6.2314,
      "step": 264500
    },
    {
      "epoch": 3.62,
      "learning_rate": 2.8405980206359235e-06,
      "loss": 6.2312,
      "step": 265000
    },
    {
      "epoch": 3.63,
      "learning_rate": 2.826559977539131e-06,
      "loss": 6.2278,
      "step": 265500
    },
    {
      "epoch": 3.63,
      "learning_rate": 2.8125219344423387e-06,
      "loss": 6.2259,
      "step": 266000
    },
    {
      "epoch": 3.64,
      "learning_rate": 2.7984838913455463e-06,
      "loss": 6.2301,
      "step": 266500
    },
    {
      "epoch": 3.65,
      "learning_rate": 2.784445848248754e-06,
      "loss": 6.2305,
      "step": 267000
    },
    {
      "epoch": 3.65,
      "learning_rate": 2.7704078051519624e-06,
      "loss": 6.2225,
      "step": 267500
    },
    {
      "epoch": 3.66,
      "learning_rate": 2.75636976205517e-06,
      "loss": 6.229,
      "step": 268000
    },
    {
      "epoch": 3.67,
      "learning_rate": 2.7423317189583776e-06,
      "loss": 6.2269,
      "step": 268500
    },
    {
      "epoch": 3.67,
      "learning_rate": 2.728293675861585e-06,
      "loss": 6.2297,
      "step": 269000
    },
    {
      "epoch": 3.68,
      "learning_rate": 2.714255632764793e-06,
      "loss": 6.229,
      "step": 269500
    },
    {
      "epoch": 3.69,
      "learning_rate": 2.7002175896680004e-06,
      "loss": 6.2186,
      "step": 270000
    },
    {
      "epoch": 3.69,
      "eval_accuracy": 0.14291574295548684,
      "eval_loss": 6.224853515625,
      "eval_runtime": 593.0153,
      "eval_samples_per_second": 519.98,
      "eval_steps_per_second": 6.501,
      "step": 270000
    },
    {
      "epoch": 3.69,
      "learning_rate": 2.686179546571208e-06,
      "loss": 6.2296,
      "step": 270500
    },
    {
      "epoch": 3.7,
      "learning_rate": 2.6721415034744156e-06,
      "loss": 6.2292,
      "step": 271000
    },
    {
      "epoch": 3.71,
      "learning_rate": 2.6581034603776237e-06,
      "loss": 6.2271,
      "step": 271500
    },
    {
      "epoch": 3.71,
      "learning_rate": 2.6440654172808313e-06,
      "loss": 6.226,
      "step": 272000
    },
    {
      "epoch": 3.72,
      "learning_rate": 2.630027374184039e-06,
      "loss": 6.2268,
      "step": 272500
    },
    {
      "epoch": 3.73,
      "learning_rate": 2.6159893310872465e-06,
      "loss": 6.2279,
      "step": 273000
    },
    {
      "epoch": 3.73,
      "learning_rate": 2.6019512879904545e-06,
      "loss": 6.2196,
      "step": 273500
    },
    {
      "epoch": 3.74,
      "learning_rate": 2.587913244893662e-06,
      "loss": 6.218,
      "step": 274000
    },
    {
      "epoch": 3.75,
      "learning_rate": 2.5738752017968697e-06,
      "loss": 6.2266,
      "step": 274500
    },
    {
      "epoch": 3.76,
      "learning_rate": 2.5598371587000773e-06,
      "loss": 6.2294,
      "step": 275000
    },
    {
      "epoch": 3.76,
      "learning_rate": 2.5457991156032854e-06,
      "loss": 6.2206,
      "step": 275500
    },
    {
      "epoch": 3.77,
      "learning_rate": 2.531761072506493e-06,
      "loss": 6.2277,
      "step": 276000
    },
    {
      "epoch": 3.78,
      "learning_rate": 2.5177230294097006e-06,
      "loss": 6.2248,
      "step": 276500
    },
    {
      "epoch": 3.78,
      "learning_rate": 2.503684986312908e-06,
      "loss": 6.2309,
      "step": 277000
    },
    {
      "epoch": 3.79,
      "learning_rate": 2.4896469432161158e-06,
      "loss": 6.2207,
      "step": 277500
    },
    {
      "epoch": 3.8,
      "learning_rate": 2.475608900119324e-06,
      "loss": 6.2243,
      "step": 278000
    },
    {
      "epoch": 3.8,
      "learning_rate": 2.4615708570225314e-06,
      "loss": 6.2216,
      "step": 278500
    },
    {
      "epoch": 3.81,
      "learning_rate": 2.447532813925739e-06,
      "loss": 6.2191,
      "step": 279000
    },
    {
      "epoch": 3.82,
      "learning_rate": 2.4334947708289466e-06,
      "loss": 6.2245,
      "step": 279500
    },
    {
      "epoch": 3.82,
      "learning_rate": 2.4194567277321542e-06,
      "loss": 6.2192,
      "step": 280000
    },
    {
      "epoch": 3.82,
      "eval_accuracy": 0.14283641610442016,
      "eval_loss": 6.223134994506836,
      "eval_runtime": 593.5494,
      "eval_samples_per_second": 519.512,
      "eval_steps_per_second": 6.495,
      "step": 280000
    },
    {
      "epoch": 3.83,
      "learning_rate": 2.405418684635362e-06,
      "loss": 6.2181,
      "step": 280500
    },
    {
      "epoch": 3.84,
      "learning_rate": 2.3913806415385694e-06,
      "loss": 6.2212,
      "step": 281000
    },
    {
      "epoch": 3.84,
      "learning_rate": 2.3773425984417775e-06,
      "loss": 6.2195,
      "step": 281500
    },
    {
      "epoch": 3.85,
      "learning_rate": 2.363304555344985e-06,
      "loss": 6.2225,
      "step": 282000
    },
    {
      "epoch": 3.86,
      "learning_rate": 2.3492665122481927e-06,
      "loss": 6.2203,
      "step": 282500
    },
    {
      "epoch": 3.86,
      "learning_rate": 2.3352284691514003e-06,
      "loss": 6.2232,
      "step": 283000
    },
    {
      "epoch": 3.87,
      "learning_rate": 2.321190426054608e-06,
      "loss": 6.2266,
      "step": 283500
    },
    {
      "epoch": 3.88,
      "learning_rate": 2.307152382957816e-06,
      "loss": 6.2182,
      "step": 284000
    },
    {
      "epoch": 3.88,
      "learning_rate": 2.2931143398610235e-06,
      "loss": 6.2208,
      "step": 284500
    },
    {
      "epoch": 3.89,
      "learning_rate": 2.279076296764231e-06,
      "loss": 6.2258,
      "step": 285000
    },
    {
      "epoch": 3.9,
      "learning_rate": 2.2650382536674387e-06,
      "loss": 6.219,
      "step": 285500
    },
    {
      "epoch": 3.91,
      "learning_rate": 2.2510002105706468e-06,
      "loss": 6.2216,
      "step": 286000
    },
    {
      "epoch": 3.91,
      "learning_rate": 2.2369621674738544e-06,
      "loss": 6.2224,
      "step": 286500
    },
    {
      "epoch": 3.92,
      "learning_rate": 2.222924124377062e-06,
      "loss": 6.2249,
      "step": 287000
    },
    {
      "epoch": 3.93,
      "learning_rate": 2.20888608128027e-06,
      "loss": 6.2227,
      "step": 287500
    },
    {
      "epoch": 3.93,
      "learning_rate": 2.1948480381834776e-06,
      "loss": 6.2224,
      "step": 288000
    },
    {
      "epoch": 3.94,
      "learning_rate": 2.1808099950866852e-06,
      "loss": 6.2214,
      "step": 288500
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.166771951989893e-06,
      "loss": 6.2214,
      "step": 289000
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.1527339088931004e-06,
      "loss": 6.2187,
      "step": 289500
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.138695865796308e-06,
      "loss": 6.2228,
      "step": 290000
    },
    {
      "epoch": 3.96,
      "eval_accuracy": 0.14331804407994586,
      "eval_loss": 6.220794200897217,
      "eval_runtime": 593.6226,
      "eval_samples_per_second": 519.448,
      "eval_steps_per_second": 6.494,
      "step": 290000
    },
    {
      "epoch": 3.97,
      "learning_rate": 2.124657822699516e-06,
      "loss": 6.22,
      "step": 290500
    },
    {
      "epoch": 3.97,
      "learning_rate": 2.1106197796027237e-06,
      "loss": 6.2202,
      "step": 291000
    },
    {
      "epoch": 3.98,
      "learning_rate": 2.0965817365059313e-06,
      "loss": 6.2174,
      "step": 291500
    },
    {
      "epoch": 3.99,
      "learning_rate": 2.082543693409139e-06,
      "loss": 6.2239,
      "step": 292000
    },
    {
      "epoch": 3.99,
      "learning_rate": 2.0685056503123465e-06,
      "loss": 6.2194,
      "step": 292500
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.054467607215554e-06,
      "loss": 6.2275,
      "step": 293000
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.040429564118762e-06,
      "loss": 6.2227,
      "step": 293500
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.0263915210219697e-06,
      "loss": 6.2223,
      "step": 294000
    },
    {
      "epoch": 4.02,
      "learning_rate": 2.0123534779251774e-06,
      "loss": 6.2263,
      "step": 294500
    },
    {
      "epoch": 4.03,
      "learning_rate": 1.998315434828385e-06,
      "loss": 6.2156,
      "step": 295000
    },
    {
      "epoch": 4.03,
      "learning_rate": 1.9842773917315926e-06,
      "loss": 6.2224,
      "step": 295500
    },
    {
      "epoch": 4.04,
      "learning_rate": 1.9702393486348e-06,
      "loss": 6.2205,
      "step": 296000
    },
    {
      "epoch": 4.05,
      "learning_rate": 1.956201305538008e-06,
      "loss": 6.2239,
      "step": 296500
    },
    {
      "epoch": 4.06,
      "learning_rate": 1.942163262441216e-06,
      "loss": 6.2187,
      "step": 297000
    },
    {
      "epoch": 4.06,
      "learning_rate": 1.9281252193444234e-06,
      "loss": 6.2191,
      "step": 297500
    },
    {
      "epoch": 4.07,
      "learning_rate": 1.9140871762476314e-06,
      "loss": 6.2216,
      "step": 298000
    },
    {
      "epoch": 4.08,
      "learning_rate": 1.9000491331508388e-06,
      "loss": 6.2194,
      "step": 298500
    },
    {
      "epoch": 4.08,
      "learning_rate": 1.8860110900540464e-06,
      "loss": 6.2153,
      "step": 299000
    },
    {
      "epoch": 4.09,
      "learning_rate": 1.8719730469572543e-06,
      "loss": 6.2187,
      "step": 299500
    },
    {
      "epoch": 4.1,
      "learning_rate": 1.857935003860462e-06,
      "loss": 6.2154,
      "step": 300000
    },
    {
      "epoch": 4.1,
      "eval_accuracy": 0.14339310791794782,
      "eval_loss": 6.2191853523254395,
      "eval_runtime": 596.5015,
      "eval_samples_per_second": 516.941,
      "eval_steps_per_second": 6.463,
      "step": 300000
    },
    {
      "epoch": 4.1,
      "learning_rate": 1.8438969607636697e-06,
      "loss": 6.2188,
      "step": 300500
    },
    {
      "epoch": 4.11,
      "learning_rate": 1.8298589176668775e-06,
      "loss": 6.2208,
      "step": 301000
    },
    {
      "epoch": 4.12,
      "learning_rate": 1.8158208745700851e-06,
      "loss": 6.2148,
      "step": 301500
    },
    {
      "epoch": 4.12,
      "learning_rate": 1.8017828314732927e-06,
      "loss": 6.2177,
      "step": 302000
    },
    {
      "epoch": 4.13,
      "learning_rate": 1.7877447883765003e-06,
      "loss": 6.2158,
      "step": 302500
    },
    {
      "epoch": 4.14,
      "learning_rate": 1.7737067452797083e-06,
      "loss": 6.2132,
      "step": 303000
    },
    {
      "epoch": 4.14,
      "learning_rate": 1.759668702182916e-06,
      "loss": 6.2183,
      "step": 303500
    },
    {
      "epoch": 4.15,
      "learning_rate": 1.7456306590861236e-06,
      "loss": 6.2185,
      "step": 304000
    },
    {
      "epoch": 4.16,
      "learning_rate": 1.7315926159893312e-06,
      "loss": 6.2215,
      "step": 304500
    },
    {
      "epoch": 4.16,
      "learning_rate": 1.7175545728925388e-06,
      "loss": 6.2254,
      "step": 305000
    },
    {
      "epoch": 4.17,
      "learning_rate": 1.7035165297957466e-06,
      "loss": 6.2227,
      "step": 305500
    },
    {
      "epoch": 4.18,
      "learning_rate": 1.6894784866989544e-06,
      "loss": 6.2141,
      "step": 306000
    },
    {
      "epoch": 4.19,
      "learning_rate": 1.675440443602162e-06,
      "loss": 6.2135,
      "step": 306500
    },
    {
      "epoch": 4.19,
      "learning_rate": 1.6614024005053696e-06,
      "loss": 6.2141,
      "step": 307000
    },
    {
      "epoch": 4.2,
      "learning_rate": 1.6473643574085774e-06,
      "loss": 6.2153,
      "step": 307500
    },
    {
      "epoch": 4.21,
      "learning_rate": 1.633326314311785e-06,
      "loss": 6.2161,
      "step": 308000
    },
    {
      "epoch": 4.21,
      "learning_rate": 1.6192882712149926e-06,
      "loss": 6.2176,
      "step": 308500
    },
    {
      "epoch": 4.22,
      "learning_rate": 1.6052502281182007e-06,
      "loss": 6.213,
      "step": 309000
    },
    {
      "epoch": 4.23,
      "learning_rate": 1.5912121850214083e-06,
      "loss": 6.2173,
      "step": 309500
    },
    {
      "epoch": 4.23,
      "learning_rate": 1.5771741419246159e-06,
      "loss": 6.2204,
      "step": 310000
    },
    {
      "epoch": 4.23,
      "eval_accuracy": 0.1433273151608806,
      "eval_loss": 6.21768045425415,
      "eval_runtime": 597.3821,
      "eval_samples_per_second": 516.179,
      "eval_steps_per_second": 6.453,
      "step": 310000
    },
    {
      "epoch": 4.24,
      "learning_rate": 1.5631360988278235e-06,
      "loss": 6.2198,
      "step": 310500
    },
    {
      "epoch": 4.25,
      "learning_rate": 1.549098055731031e-06,
      "loss": 6.2192,
      "step": 311000
    },
    {
      "epoch": 4.25,
      "learning_rate": 1.535060012634239e-06,
      "loss": 6.2156,
      "step": 311500
    },
    {
      "epoch": 4.26,
      "learning_rate": 1.5210219695374467e-06,
      "loss": 6.2163,
      "step": 312000
    },
    {
      "epoch": 4.27,
      "learning_rate": 1.5069839264406543e-06,
      "loss": 6.2133,
      "step": 312500
    },
    {
      "epoch": 4.27,
      "learning_rate": 1.492945883343862e-06,
      "loss": 6.2121,
      "step": 313000
    },
    {
      "epoch": 4.28,
      "learning_rate": 1.4789078402470698e-06,
      "loss": 6.217,
      "step": 313500
    },
    {
      "epoch": 4.29,
      "learning_rate": 1.4648697971502774e-06,
      "loss": 6.2185,
      "step": 314000
    },
    {
      "epoch": 4.29,
      "learning_rate": 1.450831754053485e-06,
      "loss": 6.2107,
      "step": 314500
    },
    {
      "epoch": 4.3,
      "learning_rate": 1.4367937109566926e-06,
      "loss": 6.2126,
      "step": 315000
    },
    {
      "epoch": 4.31,
      "learning_rate": 1.4227556678599006e-06,
      "loss": 6.2154,
      "step": 315500
    },
    {
      "epoch": 4.31,
      "learning_rate": 1.4087176247631082e-06,
      "loss": 6.2133,
      "step": 316000
    },
    {
      "epoch": 4.32,
      "learning_rate": 1.3946795816663158e-06,
      "loss": 6.2165,
      "step": 316500
    },
    {
      "epoch": 4.33,
      "learning_rate": 1.3806415385695234e-06,
      "loss": 6.2156,
      "step": 317000
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.366603495472731e-06,
      "loss": 6.2181,
      "step": 317500
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.3525654523759389e-06,
      "loss": 6.2191,
      "step": 318000
    },
    {
      "epoch": 4.35,
      "learning_rate": 1.3385274092791467e-06,
      "loss": 6.2116,
      "step": 318500
    },
    {
      "epoch": 4.36,
      "learning_rate": 1.3244893661823543e-06,
      "loss": 6.2135,
      "step": 319000
    },
    {
      "epoch": 4.36,
      "learning_rate": 1.310451323085562e-06,
      "loss": 6.2208,
      "step": 319500
    },
    {
      "epoch": 4.37,
      "learning_rate": 1.2964132799887697e-06,
      "loss": 6.2123,
      "step": 320000
    },
    {
      "epoch": 4.37,
      "eval_accuracy": 0.1433558993505167,
      "eval_loss": 6.216210842132568,
      "eval_runtime": 596.5266,
      "eval_samples_per_second": 516.919,
      "eval_steps_per_second": 6.462,
      "step": 320000
    },
    {
      "epoch": 4.38,
      "learning_rate": 1.2823752368919773e-06,
      "loss": 6.2155,
      "step": 320500
    },
    {
      "epoch": 4.38,
      "learning_rate": 1.268337193795185e-06,
      "loss": 6.2211,
      "step": 321000
    },
    {
      "epoch": 4.39,
      "learning_rate": 1.254299150698393e-06,
      "loss": 6.2143,
      "step": 321500
    },
    {
      "epoch": 4.4,
      "learning_rate": 1.2402611076016003e-06,
      "loss": 6.2151,
      "step": 322000
    },
    {
      "epoch": 4.4,
      "learning_rate": 1.2262230645048082e-06,
      "loss": 6.2124,
      "step": 322500
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.2121850214080158e-06,
      "loss": 6.2164,
      "step": 323000
    },
    {
      "epoch": 4.42,
      "learning_rate": 1.1981469783112234e-06,
      "loss": 6.2162,
      "step": 323500
    },
    {
      "epoch": 4.42,
      "learning_rate": 1.1841089352144312e-06,
      "loss": 6.2163,
      "step": 324000
    },
    {
      "epoch": 4.43,
      "learning_rate": 1.1700708921176388e-06,
      "loss": 6.2157,
      "step": 324500
    },
    {
      "epoch": 4.44,
      "learning_rate": 1.1560328490208466e-06,
      "loss": 6.2169,
      "step": 325000
    },
    {
      "epoch": 4.44,
      "learning_rate": 1.1419948059240542e-06,
      "loss": 6.216,
      "step": 325500
    },
    {
      "epoch": 4.45,
      "learning_rate": 1.127956762827262e-06,
      "loss": 6.2143,
      "step": 326000
    },
    {
      "epoch": 4.46,
      "learning_rate": 1.1139187197304696e-06,
      "loss": 6.2161,
      "step": 326500
    },
    {
      "epoch": 4.47,
      "learning_rate": 1.0998806766336775e-06,
      "loss": 6.2144,
      "step": 327000
    },
    {
      "epoch": 4.47,
      "learning_rate": 1.085842633536885e-06,
      "loss": 6.2149,
      "step": 327500
    },
    {
      "epoch": 4.48,
      "learning_rate": 1.0718045904400927e-06,
      "loss": 6.2202,
      "step": 328000
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.0577665473433005e-06,
      "loss": 6.2104,
      "step": 328500
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.043728504246508e-06,
      "loss": 6.2138,
      "step": 329000
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.0296904611497157e-06,
      "loss": 6.2125,
      "step": 329500
    },
    {
      "epoch": 4.51,
      "learning_rate": 1.0156524180529235e-06,
      "loss": 6.2136,
      "step": 330000
    },
    {
      "epoch": 4.51,
      "eval_accuracy": 0.14372247288394874,
      "eval_loss": 6.213418960571289,
      "eval_runtime": 595.2765,
      "eval_samples_per_second": 518.005,
      "eval_steps_per_second": 6.476,
      "step": 330000
    },
    {
      "epoch": 4.51,
      "learning_rate": 1.0016143749561311e-06,
      "loss": 6.2106,
      "step": 330500
    },
    {
      "epoch": 4.52,
      "learning_rate": 9.87576331859339e-07,
      "loss": 6.2167,
      "step": 331000
    },
    {
      "epoch": 4.53,
      "learning_rate": 9.735382887625466e-07,
      "loss": 6.2154,
      "step": 331500
    },
    {
      "epoch": 4.53,
      "learning_rate": 9.595002456657544e-07,
      "loss": 6.2211,
      "step": 332000
    },
    {
      "epoch": 4.54,
      "learning_rate": 9.45462202568962e-07,
      "loss": 6.2151,
      "step": 332500
    },
    {
      "epoch": 4.55,
      "learning_rate": 9.314241594721697e-07,
      "loss": 6.2177,
      "step": 333000
    },
    {
      "epoch": 4.55,
      "learning_rate": 9.173861163753774e-07,
      "loss": 6.2124,
      "step": 333500
    },
    {
      "epoch": 4.56,
      "learning_rate": 9.03348073278585e-07,
      "loss": 6.2149,
      "step": 334000
    },
    {
      "epoch": 4.57,
      "learning_rate": 8.893100301817928e-07,
      "loss": 6.2102,
      "step": 334500
    },
    {
      "epoch": 4.57,
      "learning_rate": 8.752719870850004e-07,
      "loss": 6.2175,
      "step": 335000
    },
    {
      "epoch": 4.58,
      "learning_rate": 8.612339439882081e-07,
      "loss": 6.2157,
      "step": 335500
    },
    {
      "epoch": 4.59,
      "learning_rate": 8.471959008914159e-07,
      "loss": 6.2142,
      "step": 336000
    },
    {
      "epoch": 4.59,
      "learning_rate": 8.331578577946236e-07,
      "loss": 6.2156,
      "step": 336500
    },
    {
      "epoch": 4.6,
      "learning_rate": 8.191198146978312e-07,
      "loss": 6.2152,
      "step": 337000
    },
    {
      "epoch": 4.61,
      "learning_rate": 8.050817716010388e-07,
      "loss": 6.216,
      "step": 337500
    },
    {
      "epoch": 4.62,
      "learning_rate": 7.910437285042466e-07,
      "loss": 6.2122,
      "step": 338000
    },
    {
      "epoch": 4.62,
      "learning_rate": 7.770056854074542e-07,
      "loss": 6.2142,
      "step": 338500
    },
    {
      "epoch": 4.63,
      "learning_rate": 7.629676423106619e-07,
      "loss": 6.2145,
      "step": 339000
    },
    {
      "epoch": 4.64,
      "learning_rate": 7.489295992138697e-07,
      "loss": 6.2139,
      "step": 339500
    },
    {
      "epoch": 4.64,
      "learning_rate": 7.348915561170773e-07,
      "loss": 6.214,
      "step": 340000
    },
    {
      "epoch": 4.64,
      "eval_accuracy": 0.1438502864694206,
      "eval_loss": 6.213069438934326,
      "eval_runtime": 592.9517,
      "eval_samples_per_second": 520.036,
      "eval_steps_per_second": 6.501,
      "step": 340000
    },
    {
      "epoch": 4.65,
      "learning_rate": 7.208535130202849e-07,
      "loss": 6.2128,
      "step": 340500
    },
    {
      "epoch": 4.66,
      "learning_rate": 7.068154699234928e-07,
      "loss": 6.2134,
      "step": 341000
    },
    {
      "epoch": 4.66,
      "learning_rate": 6.927774268267004e-07,
      "loss": 6.2088,
      "step": 341500
    },
    {
      "epoch": 4.67,
      "learning_rate": 6.787393837299081e-07,
      "loss": 6.2164,
      "step": 342000
    },
    {
      "epoch": 4.68,
      "learning_rate": 6.647013406331158e-07,
      "loss": 6.2141,
      "step": 342500
    },
    {
      "epoch": 4.68,
      "learning_rate": 6.506632975363235e-07,
      "loss": 6.2151,
      "step": 343000
    },
    {
      "epoch": 4.69,
      "learning_rate": 6.366252544395311e-07,
      "loss": 6.2107,
      "step": 343500
    },
    {
      "epoch": 4.7,
      "learning_rate": 6.225872113427388e-07,
      "loss": 6.2157,
      "step": 344000
    },
    {
      "epoch": 4.7,
      "learning_rate": 6.085491682459465e-07,
      "loss": 6.2156,
      "step": 344500
    },
    {
      "epoch": 4.71,
      "learning_rate": 5.945111251491542e-07,
      "loss": 6.2197,
      "step": 345000
    },
    {
      "epoch": 4.72,
      "learning_rate": 5.80473082052362e-07,
      "loss": 6.2121,
      "step": 345500
    },
    {
      "epoch": 4.72,
      "learning_rate": 5.664350389555697e-07,
      "loss": 6.2151,
      "step": 346000
    },
    {
      "epoch": 4.73,
      "learning_rate": 5.523969958587774e-07,
      "loss": 6.2134,
      "step": 346500
    },
    {
      "epoch": 4.74,
      "learning_rate": 5.38358952761985e-07,
      "loss": 6.211,
      "step": 347000
    },
    {
      "epoch": 4.74,
      "learning_rate": 5.243209096651927e-07,
      "loss": 6.2162,
      "step": 347500
    },
    {
      "epoch": 4.75,
      "learning_rate": 5.102828665684004e-07,
      "loss": 6.2066,
      "step": 348000
    },
    {
      "epoch": 4.76,
      "learning_rate": 4.962448234716081e-07,
      "loss": 6.2134,
      "step": 348500
    },
    {
      "epoch": 4.77,
      "learning_rate": 4.822067803748158e-07,
      "loss": 6.2099,
      "step": 349000
    },
    {
      "epoch": 4.77,
      "learning_rate": 4.6816873727802344e-07,
      "loss": 6.2162,
      "step": 349500
    },
    {
      "epoch": 4.78,
      "learning_rate": 4.5413069418123115e-07,
      "loss": 6.2139,
      "step": 350000
    },
    {
      "epoch": 4.78,
      "eval_accuracy": 0.14368222749880583,
      "eval_loss": 6.212860107421875,
      "eval_runtime": 591.4247,
      "eval_samples_per_second": 521.378,
      "eval_steps_per_second": 6.518,
      "step": 350000
    },
    {
      "epoch": 4.79,
      "learning_rate": 4.4009265108443886e-07,
      "loss": 6.2176,
      "step": 350500
    },
    {
      "epoch": 4.79,
      "learning_rate": 4.260546079876465e-07,
      "loss": 6.2179,
      "step": 351000
    },
    {
      "epoch": 4.8,
      "learning_rate": 4.1201656489085423e-07,
      "loss": 6.2203,
      "step": 351500
    },
    {
      "epoch": 4.81,
      "learning_rate": 3.9797852179406194e-07,
      "loss": 6.211,
      "step": 352000
    },
    {
      "epoch": 4.81,
      "learning_rate": 3.839404786972696e-07,
      "loss": 6.218,
      "step": 352500
    },
    {
      "epoch": 4.82,
      "learning_rate": 3.699024356004773e-07,
      "loss": 6.2146,
      "step": 353000
    },
    {
      "epoch": 4.83,
      "learning_rate": 3.5586439250368503e-07,
      "loss": 6.2107,
      "step": 353500
    },
    {
      "epoch": 4.83,
      "learning_rate": 3.418263494068927e-07,
      "loss": 6.2123,
      "step": 354000
    },
    {
      "epoch": 4.84,
      "learning_rate": 3.277883063101004e-07,
      "loss": 6.216,
      "step": 354500
    },
    {
      "epoch": 4.85,
      "learning_rate": 3.137502632133081e-07,
      "loss": 6.2138,
      "step": 355000
    },
    {
      "epoch": 4.85,
      "learning_rate": 2.9971222011651577e-07,
      "loss": 6.2132,
      "step": 355500
    },
    {
      "epoch": 4.86,
      "learning_rate": 2.856741770197235e-07,
      "loss": 6.2114,
      "step": 356000
    },
    {
      "epoch": 4.87,
      "learning_rate": 2.7163613392293114e-07,
      "loss": 6.214,
      "step": 356500
    },
    {
      "epoch": 4.87,
      "learning_rate": 2.5759809082613885e-07,
      "loss": 6.2111,
      "step": 357000
    },
    {
      "epoch": 4.88,
      "learning_rate": 2.4356004772934656e-07,
      "loss": 6.2109,
      "step": 357500
    },
    {
      "epoch": 4.89,
      "learning_rate": 2.2952200463255425e-07,
      "loss": 6.2154,
      "step": 358000
    },
    {
      "epoch": 4.9,
      "learning_rate": 2.1548396153576193e-07,
      "loss": 6.2192,
      "step": 358500
    },
    {
      "epoch": 4.9,
      "learning_rate": 2.0144591843896962e-07,
      "loss": 6.2091,
      "step": 359000
    },
    {
      "epoch": 4.91,
      "learning_rate": 1.8740787534217733e-07,
      "loss": 6.2121,
      "step": 359500
    },
    {
      "epoch": 4.92,
      "learning_rate": 1.7336983224538502e-07,
      "loss": 6.211,
      "step": 360000
    },
    {
      "epoch": 4.92,
      "eval_accuracy": 0.1438829220447318,
      "eval_loss": 6.213039875030518,
      "eval_runtime": 592.7416,
      "eval_samples_per_second": 520.22,
      "eval_steps_per_second": 6.504,
      "step": 360000
    },
    {
      "epoch": 4.92,
      "learning_rate": 1.5933178914859268e-07,
      "loss": 6.2094,
      "step": 360500
    },
    {
      "epoch": 4.93,
      "learning_rate": 1.452937460518004e-07,
      "loss": 6.2112,
      "step": 361000
    },
    {
      "epoch": 4.94,
      "learning_rate": 1.3125570295500807e-07,
      "loss": 6.2137,
      "step": 361500
    },
    {
      "epoch": 4.94,
      "learning_rate": 1.1721765985821577e-07,
      "loss": 6.2146,
      "step": 362000
    },
    {
      "epoch": 4.95,
      "learning_rate": 1.0317961676142347e-07,
      "loss": 6.2102,
      "step": 362500
    },
    {
      "epoch": 4.96,
      "learning_rate": 8.914157366463117e-08,
      "loss": 6.2139,
      "step": 363000
    },
    {
      "epoch": 4.96,
      "learning_rate": 7.510353056783884e-08,
      "loss": 6.2126,
      "step": 363500
    },
    {
      "epoch": 4.97,
      "learning_rate": 6.106548747104654e-08,
      "loss": 6.2099,
      "step": 364000
    },
    {
      "epoch": 4.98,
      "learning_rate": 4.702744437425423e-08,
      "loss": 6.2121,
      "step": 364500
    },
    {
      "epoch": 4.98,
      "learning_rate": 3.2989401277461925e-08,
      "loss": 6.2119,
      "step": 365000
    },
    {
      "epoch": 4.99,
      "learning_rate": 1.8951358180669617e-08,
      "loss": 6.2124,
      "step": 365500
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.913315083877309e-09,
      "loss": 6.2139,
      "step": 366000
    },
    {
      "epoch": 5.0,
      "step": 366175,
      "total_flos": 2.055540121440092e+17,
      "train_loss": 6.426997217993659,
      "train_runtime": 130490.9707,
      "train_samples_per_second": 224.489,
      "train_steps_per_second": 2.806
    }
  ],
  "logging_steps": 500,
  "max_steps": 366175,
  "num_train_epochs": 5,
  "save_steps": 10000,
  "total_flos": 2.055540121440092e+17,
  "trial_name": null,
  "trial_params": null
}