{
  "best_metric": 0.9895390868186951,
  "best_model_checkpoint": "/kaggle/output/checkpoint-49000",
  "epoch": 2.1186440677966103,
  "eval_steps": 1000,
  "global_step": 52000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2.7777777777777777e-11,
      "loss": 1.2076,
      "step": 1
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.7750000000000004e-08,
      "loss": 1.1301,
      "step": 1000
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 0.3321357285429142,
      "eval_loss": 1.1047232151031494,
      "eval_runtime": 50.8344,
      "eval_samples_per_second": 98.555,
      "eval_steps_per_second": 12.334,
      "step": 1000
    },
    {
      "epoch": 0.08,
      "learning_rate": 5.5527777777777784e-08,
      "loss": 1.1286,
      "step": 2000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.3377245508982036,
      "eval_loss": 1.1038333177566528,
      "eval_runtime": 50.8114,
      "eval_samples_per_second": 98.6,
      "eval_steps_per_second": 12.34,
      "step": 2000
    },
    {
      "epoch": 0.12,
      "learning_rate": 8.327777777777778e-08,
      "loss": 1.1214,
      "step": 3000
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.3409181636726547,
      "eval_loss": 1.103529930114746,
      "eval_runtime": 50.8258,
      "eval_samples_per_second": 98.572,
      "eval_steps_per_second": 12.336,
      "step": 3000
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.1105555555555557e-07,
      "loss": 1.1249,
      "step": 4000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.3469061876247505,
      "eval_loss": 1.1000434160232544,
      "eval_runtime": 50.6414,
      "eval_samples_per_second": 98.931,
      "eval_steps_per_second": 12.381,
      "step": 4000
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.3880555555555558e-07,
      "loss": 1.1206,
      "step": 5000
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.34171656686626745,
      "eval_loss": 1.0996097326278687,
      "eval_runtime": 50.6956,
      "eval_samples_per_second": 98.825,
      "eval_steps_per_second": 12.368,
      "step": 5000
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.6658333333333335e-07,
      "loss": 1.1161,
      "step": 6000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.3590818363273453,
      "eval_loss": 1.096977710723877,
      "eval_runtime": 50.5826,
      "eval_samples_per_second": 99.046,
      "eval_steps_per_second": 12.396,
      "step": 6000
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9433333333333334e-07,
      "loss": 1.1138,
      "step": 7000
    },
    {
      "epoch": 0.29,
      "eval_accuracy": 0.3532934131736527,
      "eval_loss": 1.0949815511703491,
      "eval_runtime": 50.5604,
      "eval_samples_per_second": 99.089,
      "eval_steps_per_second": 12.401,
      "step": 7000
    },
    {
      "epoch": 0.33,
      "learning_rate": 2.2211111111111114e-07,
      "loss": 1.1135,
      "step": 8000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.36327345309381237,
      "eval_loss": 1.093558430671692,
      "eval_runtime": 50.608,
      "eval_samples_per_second": 98.996,
      "eval_steps_per_second": 12.389,
      "step": 8000
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.4986111111111113e-07,
      "loss": 1.1063,
      "step": 9000
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.35548902195608784,
      "eval_loss": 1.0992107391357422,
      "eval_runtime": 50.5434,
      "eval_samples_per_second": 99.123,
      "eval_steps_per_second": 12.405,
      "step": 9000
    },
    {
      "epoch": 0.41,
      "learning_rate": 2.776388888888889e-07,
      "loss": 1.1083,
      "step": 10000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.3836327345309381,
      "eval_loss": 1.0920206308364868,
      "eval_runtime": 50.5559,
      "eval_samples_per_second": 99.098,
      "eval_steps_per_second": 12.402,
      "step": 10000
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.053888888888889e-07,
      "loss": 1.1044,
      "step": 11000
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.38762475049900197,
      "eval_loss": 1.0904459953308105,
      "eval_runtime": 50.9725,
      "eval_samples_per_second": 98.288,
      "eval_steps_per_second": 12.301,
      "step": 11000
    },
    {
      "epoch": 0.49,
      "learning_rate": 3.331666666666667e-07,
      "loss": 1.1019,
      "step": 12000
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.3912175648702595,
      "eval_loss": 1.088543176651001,
      "eval_runtime": 50.8297,
      "eval_samples_per_second": 98.564,
      "eval_steps_per_second": 12.335,
      "step": 12000
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.609166666666667e-07,
      "loss": 1.1032,
      "step": 13000
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 0.3435129740518962,
      "eval_loss": 1.0940682888031006,
      "eval_runtime": 51.0031,
      "eval_samples_per_second": 98.229,
      "eval_steps_per_second": 12.293,
      "step": 13000
    },
    {
      "epoch": 0.57,
      "learning_rate": 3.886944444444445e-07,
      "loss": 1.1006,
      "step": 14000
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.36826347305389223,
      "eval_loss": 1.0924910306930542,
      "eval_runtime": 51.2753,
      "eval_samples_per_second": 97.708,
      "eval_steps_per_second": 12.228,
      "step": 14000
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.164444444444445e-07,
      "loss": 1.1,
      "step": 15000
    },
    {
      "epoch": 0.61,
      "eval_accuracy": 0.38323353293413176,
      "eval_loss": 1.0879395008087158,
      "eval_runtime": 51.3783,
      "eval_samples_per_second": 97.512,
      "eval_steps_per_second": 12.204,
      "step": 15000
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.442222222222223e-07,
      "loss": 1.0982,
      "step": 16000
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.3964071856287425,
      "eval_loss": 1.0822893381118774,
      "eval_runtime": 51.2346,
      "eval_samples_per_second": 97.786,
      "eval_steps_per_second": 12.238,
      "step": 16000
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.7197222222222224e-07,
      "loss": 1.0953,
      "step": 17000
    },
    {
      "epoch": 0.69,
      "eval_accuracy": 0.4121756487025948,
      "eval_loss": 1.0795228481292725,
      "eval_runtime": 51.31,
      "eval_samples_per_second": 97.642,
      "eval_steps_per_second": 12.22,
      "step": 17000
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.997500000000001e-07,
      "loss": 1.0943,
      "step": 18000
    },
    {
      "epoch": 0.73,
      "eval_accuracy": 0.42015968063872255,
      "eval_loss": 1.079458236694336,
      "eval_runtime": 51.3255,
      "eval_samples_per_second": 97.612,
      "eval_steps_per_second": 12.216,
      "step": 18000
    },
    {
      "epoch": 0.77,
      "learning_rate": 5.275e-07,
      "loss": 1.0885,
      "step": 19000
    },
    {
      "epoch": 0.77,
      "eval_accuracy": 0.437125748502994,
      "eval_loss": 1.0667964220046997,
      "eval_runtime": 51.1737,
      "eval_samples_per_second": 97.902,
      "eval_steps_per_second": 12.252,
      "step": 19000
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.552777777777778e-07,
      "loss": 1.0792,
      "step": 20000
    },
    {
      "epoch": 0.81,
      "eval_accuracy": 0.4315369261477046,
      "eval_loss": 1.0628777742385864,
      "eval_runtime": 51.4162,
      "eval_samples_per_second": 97.44,
      "eval_steps_per_second": 12.195,
      "step": 20000
    },
    {
      "epoch": 0.86,
      "learning_rate": 5.830277777777779e-07,
      "loss": 1.0755,
      "step": 21000
    },
    {
      "epoch": 0.86,
      "eval_accuracy": 0.4343313373253493,
      "eval_loss": 1.0604455471038818,
      "eval_runtime": 51.2629,
      "eval_samples_per_second": 97.731,
      "eval_steps_per_second": 12.231,
      "step": 21000
    },
    {
      "epoch": 0.9,
      "learning_rate": 6.108055555555556e-07,
      "loss": 1.073,
      "step": 22000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.43353293413173655,
      "eval_loss": 1.0552711486816406,
      "eval_runtime": 51.2402,
      "eval_samples_per_second": 97.775,
      "eval_steps_per_second": 12.236,
      "step": 22000
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.385555555555556e-07,
      "loss": 1.0699,
      "step": 23000
    },
    {
      "epoch": 0.94,
      "eval_accuracy": 0.4313373253493014,
      "eval_loss": 1.052398920059204,
      "eval_runtime": 51.1758,
      "eval_samples_per_second": 97.898,
      "eval_steps_per_second": 12.252,
      "step": 23000
    },
    {
      "epoch": 0.98,
      "learning_rate": 6.663333333333334e-07,
      "loss": 1.0695,
      "step": 24000
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.4407185628742515,
      "eval_loss": 1.047132968902588,
      "eval_runtime": 51.3549,
      "eval_samples_per_second": 97.556,
      "eval_steps_per_second": 12.209,
      "step": 24000
    },
    {
      "epoch": 1.02,
      "learning_rate": 6.940833333333334e-07,
      "loss": 1.0656,
      "step": 25000
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.4439121756487026,
      "eval_loss": 1.045853853225708,
      "eval_runtime": 51.1548,
      "eval_samples_per_second": 97.938,
      "eval_steps_per_second": 12.257,
      "step": 25000
    },
    {
      "epoch": 1.06,
      "learning_rate": 7.218611111111112e-07,
      "loss": 1.0645,
      "step": 26000
    },
    {
      "epoch": 1.06,
      "eval_accuracy": 0.43273453093812375,
      "eval_loss": 1.0467982292175293,
      "eval_runtime": 51.1774,
      "eval_samples_per_second": 97.895,
      "eval_steps_per_second": 12.252,
      "step": 26000
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.496111111111112e-07,
      "loss": 1.0639,
      "step": 27000
    },
    {
      "epoch": 1.1,
      "eval_accuracy": 0.43313373253493015,
      "eval_loss": 1.0436949729919434,
      "eval_runtime": 51.1313,
      "eval_samples_per_second": 97.983,
      "eval_steps_per_second": 12.263,
      "step": 27000
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.77388888888889e-07,
      "loss": 1.0638,
      "step": 28000
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.436127744510978,
      "eval_loss": 1.0424726009368896,
      "eval_runtime": 51.2359,
      "eval_samples_per_second": 97.783,
      "eval_steps_per_second": 12.238,
      "step": 28000
    },
    {
      "epoch": 1.18,
      "learning_rate": 8.051388888888889e-07,
      "loss": 1.0591,
      "step": 29000
    },
    {
      "epoch": 1.18,
      "eval_accuracy": 0.43852295409181635,
      "eval_loss": 1.0379680395126343,
      "eval_runtime": 51.0917,
      "eval_samples_per_second": 98.059,
      "eval_steps_per_second": 12.272,
      "step": 29000
    },
    {
      "epoch": 1.22,
      "learning_rate": 8.329166666666667e-07,
      "loss": 1.0571,
      "step": 30000
    },
    {
      "epoch": 1.22,
      "eval_accuracy": 0.43932135728542915,
      "eval_loss": 1.034009575843811,
      "eval_runtime": 51.187,
      "eval_samples_per_second": 97.876,
      "eval_steps_per_second": 12.249,
      "step": 30000
    },
    {
      "epoch": 1.26,
      "learning_rate": 8.606666666666668e-07,
      "loss": 1.0609,
      "step": 31000
    },
    {
      "epoch": 1.26,
      "eval_accuracy": 0.4489021956087824,
      "eval_loss": 1.0380228757858276,
      "eval_runtime": 51.2533,
      "eval_samples_per_second": 97.75,
      "eval_steps_per_second": 12.233,
      "step": 31000
    },
    {
      "epoch": 1.3,
      "learning_rate": 8.884444444444445e-07,
      "loss": 1.0536,
      "step": 32000
    },
    {
      "epoch": 1.3,
      "eval_accuracy": 0.4347305389221557,
      "eval_loss": 1.035749912261963,
      "eval_runtime": 51.1708,
      "eval_samples_per_second": 97.907,
      "eval_steps_per_second": 12.253,
      "step": 32000
    },
    {
      "epoch": 1.34,
      "learning_rate": 9.161944444444445e-07,
      "loss": 1.0526,
      "step": 33000
    },
    {
      "epoch": 1.34,
      "eval_accuracy": 0.4471057884231537,
      "eval_loss": 1.0298662185668945,
      "eval_runtime": 51.2742,
      "eval_samples_per_second": 97.71,
      "eval_steps_per_second": 12.228,
      "step": 33000
    },
    {
      "epoch": 1.39,
      "learning_rate": 9.439722222222223e-07,
      "loss": 1.0454,
      "step": 34000
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.449500998003992,
      "eval_loss": 1.028323769569397,
      "eval_runtime": 51.2636,
      "eval_samples_per_second": 97.73,
      "eval_steps_per_second": 12.231,
      "step": 34000
    },
    {
      "epoch": 1.43,
      "learning_rate": 9.717222222222224e-07,
      "loss": 1.0507,
      "step": 35000
    },
    {
      "epoch": 1.43,
      "eval_accuracy": 0.4554890219560878,
      "eval_loss": 1.0231916904449463,
      "eval_runtime": 51.2328,
      "eval_samples_per_second": 97.789,
      "eval_steps_per_second": 12.238,
      "step": 35000
    },
    {
      "epoch": 1.47,
      "learning_rate": 9.995000000000001e-07,
      "loss": 1.0462,
      "step": 36000
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.4590818363273453,
      "eval_loss": 1.02061927318573,
      "eval_runtime": 50.6455,
      "eval_samples_per_second": 98.923,
      "eval_steps_per_second": 12.38,
      "step": 36000
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.02725e-06,
      "loss": 1.0423,
      "step": 37000
    },
    {
      "epoch": 1.51,
      "eval_accuracy": 0.4722554890219561,
      "eval_loss": 1.0115904808044434,
      "eval_runtime": 51.0173,
      "eval_samples_per_second": 98.202,
      "eval_steps_per_second": 12.29,
      "step": 37000
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.055027777777778e-06,
      "loss": 1.041,
      "step": 38000
    },
    {
      "epoch": 1.55,
      "eval_accuracy": 0.47365269461077847,
      "eval_loss": 1.0123876333236694,
      "eval_runtime": 51.1777,
      "eval_samples_per_second": 97.894,
      "eval_steps_per_second": 12.251,
      "step": 38000
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.0827777777777778e-06,
      "loss": 1.0422,
      "step": 39000
    },
    {
      "epoch": 1.59,
      "eval_accuracy": 0.46367265469061875,
      "eval_loss": 1.019059181213379,
      "eval_runtime": 51.1786,
      "eval_samples_per_second": 97.892,
      "eval_steps_per_second": 12.251,
      "step": 39000
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.1105555555555556e-06,
      "loss": 1.034,
      "step": 40000
    },
    {
      "epoch": 1.63,
      "eval_accuracy": 0.4782435129740519,
      "eval_loss": 1.0045363903045654,
      "eval_runtime": 51.2481,
      "eval_samples_per_second": 97.76,
      "eval_steps_per_second": 12.235,
      "step": 40000
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.1383055555555557e-06,
      "loss": 1.0349,
      "step": 41000
    },
    {
      "epoch": 1.67,
      "eval_accuracy": 0.4550898203592814,
      "eval_loss": 1.0253337621688843,
      "eval_runtime": 51.219,
      "eval_samples_per_second": 97.815,
      "eval_steps_per_second": 12.242,
      "step": 41000
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.1660833333333334e-06,
      "loss": 1.0338,
      "step": 42000
    },
    {
      "epoch": 1.71,
      "eval_accuracy": 0.480439121756487,
      "eval_loss": 1.000636339187622,
      "eval_runtime": 51.1853,
      "eval_samples_per_second": 97.88,
      "eval_steps_per_second": 12.25,
      "step": 42000
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.1938333333333335e-06,
      "loss": 1.031,
      "step": 43000
    },
    {
      "epoch": 1.75,
      "eval_accuracy": 0.45888223552894214,
      "eval_loss": 1.0288445949554443,
      "eval_runtime": 51.295,
      "eval_samples_per_second": 97.67,
      "eval_steps_per_second": 12.223,
      "step": 43000
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.2216111111111113e-06,
      "loss": 1.0327,
      "step": 44000
    },
    {
      "epoch": 1.79,
      "eval_accuracy": 0.493812375249501,
      "eval_loss": 0.9989358186721802,
      "eval_runtime": 51.1981,
      "eval_samples_per_second": 97.855,
      "eval_steps_per_second": 12.247,
      "step": 44000
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.2493611111111112e-06,
      "loss": 1.0321,
      "step": 45000
    },
    {
      "epoch": 1.83,
      "eval_accuracy": 0.4870259481037924,
      "eval_loss": 1.003555417060852,
      "eval_runtime": 50.8148,
      "eval_samples_per_second": 98.593,
      "eval_steps_per_second": 12.339,
      "step": 45000
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.277138888888889e-06,
      "loss": 1.0326,
      "step": 46000
    },
    {
      "epoch": 1.87,
      "eval_accuracy": 0.4810379241516966,
      "eval_loss": 1.00687575340271,
      "eval_runtime": 50.7043,
      "eval_samples_per_second": 98.808,
      "eval_steps_per_second": 12.366,
      "step": 46000
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.304888888888889e-06,
      "loss": 1.0266,
      "step": 47000
    },
    {
      "epoch": 1.91,
      "eval_accuracy": 0.49840319361277446,
      "eval_loss": 0.9924121499061584,
      "eval_runtime": 50.7174,
      "eval_samples_per_second": 98.783,
      "eval_steps_per_second": 12.363,
      "step": 47000
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.3326666666666668e-06,
      "loss": 1.0285,
      "step": 48000
    },
    {
      "epoch": 1.96,
      "eval_accuracy": 0.45788423153692615,
      "eval_loss": 1.0277597904205322,
      "eval_runtime": 50.7995,
      "eval_samples_per_second": 98.623,
      "eval_steps_per_second": 12.343,
      "step": 48000
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.3604166666666668e-06,
      "loss": 1.0236,
      "step": 49000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5021956087824351,
      "eval_loss": 0.9895390868186951,
      "eval_runtime": 50.824,
      "eval_samples_per_second": 98.576,
      "eval_steps_per_second": 12.337,
      "step": 49000
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.3881944444444446e-06,
      "loss": 1.0243,
      "step": 50000
    },
    {
      "epoch": 2.04,
      "eval_accuracy": 0.47984031936127747,
      "eval_loss": 1.0068507194519043,
      "eval_runtime": 50.7665,
      "eval_samples_per_second": 98.687,
      "eval_steps_per_second": 12.351,
      "step": 50000
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.4159444444444445e-06,
      "loss": 1.0237,
      "step": 51000
    },
    {
      "epoch": 2.08,
      "eval_accuracy": 0.5053892215568863,
      "eval_loss": 0.9906230568885803,
      "eval_runtime": 50.9162,
      "eval_samples_per_second": 98.397,
      "eval_steps_per_second": 12.314,
      "step": 51000
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.4437222222222225e-06,
      "loss": 1.0184,
      "step": 52000
    },
    {
      "epoch": 2.12,
      "eval_accuracy": 0.47784431137724553,
      "eval_loss": 1.015572428703308,
      "eval_runtime": 51.1054,
      "eval_samples_per_second": 98.033,
      "eval_steps_per_second": 12.269,
      "step": 52000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 10000000,
  "num_train_epochs": 408,
  "save_steps": 1000,
  "total_flos": 1.086979441903534e+17,
  "trial_name": null,
  "trial_params": null
}