{
  "best_metric": 1.0189749002456665,
  "best_model_checkpoint": "/kaggle/output/checkpoint-38000",
  "epoch": 1.6297262059973925,
  "eval_steps": 1000,
  "global_step": 40000,
  "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": 1.1383,
      "step": 1
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.7750000000000004e-08,
      "loss": 1.1424,
      "step": 1000
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 0.32375249500998005,
      "eval_loss": 1.1077626943588257,
      "eval_runtime": 54.8633,
      "eval_samples_per_second": 91.318,
      "eval_steps_per_second": 11.428,
      "step": 1000
    },
    {
      "epoch": 0.08,
      "learning_rate": 5.5527777777777784e-08,
      "loss": 1.1244,
      "step": 2000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.33652694610778444,
      "eval_loss": 1.1080161333084106,
      "eval_runtime": 54.7384,
      "eval_samples_per_second": 91.526,
      "eval_steps_per_second": 11.454,
      "step": 2000
    },
    {
      "epoch": 0.12,
      "learning_rate": 8.327777777777778e-08,
      "loss": 1.1228,
      "step": 3000
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.34331337325349304,
      "eval_loss": 1.1084064245224,
      "eval_runtime": 54.7948,
      "eval_samples_per_second": 91.432,
      "eval_steps_per_second": 11.443,
      "step": 3000
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.1105555555555557e-07,
      "loss": 1.1216,
      "step": 4000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.3385229540918164,
      "eval_loss": 1.1014840602874756,
      "eval_runtime": 54.8508,
      "eval_samples_per_second": 91.339,
      "eval_steps_per_second": 11.431,
      "step": 4000
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.3880555555555558e-07,
      "loss": 1.1181,
      "step": 5000
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.33073852295409184,
      "eval_loss": 1.1008135080337524,
      "eval_runtime": 54.8304,
      "eval_samples_per_second": 91.373,
      "eval_steps_per_second": 11.435,
      "step": 5000
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.6658333333333335e-07,
      "loss": 1.1132,
      "step": 6000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.3520958083832335,
      "eval_loss": 1.0993762016296387,
      "eval_runtime": 54.8804,
      "eval_samples_per_second": 91.289,
      "eval_steps_per_second": 11.425,
      "step": 6000
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9433333333333334e-07,
      "loss": 1.1113,
      "step": 7000
    },
    {
      "epoch": 0.29,
      "eval_accuracy": 0.3530938123752495,
      "eval_loss": 1.0965770483016968,
      "eval_runtime": 54.8881,
      "eval_samples_per_second": 91.277,
      "eval_steps_per_second": 11.423,
      "step": 7000
    },
    {
      "epoch": 0.33,
      "learning_rate": 2.2211111111111114e-07,
      "loss": 1.1111,
      "step": 8000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.35708582834331337,
      "eval_loss": 1.094658613204956,
      "eval_runtime": 54.8233,
      "eval_samples_per_second": 91.384,
      "eval_steps_per_second": 11.437,
      "step": 8000
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.4986111111111113e-07,
      "loss": 1.109,
      "step": 9000
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.34191616766467064,
      "eval_loss": 1.106990933418274,
      "eval_runtime": 54.9095,
      "eval_samples_per_second": 91.241,
      "eval_steps_per_second": 11.419,
      "step": 9000
    },
    {
      "epoch": 0.41,
      "learning_rate": 2.776388888888889e-07,
      "loss": 1.1036,
      "step": 10000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.37584830339321357,
      "eval_loss": 1.0930211544036865,
      "eval_runtime": 54.9067,
      "eval_samples_per_second": 91.246,
      "eval_steps_per_second": 11.419,
      "step": 10000
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.0541666666666667e-07,
      "loss": 1.1045,
      "step": 11000
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.3652694610778443,
      "eval_loss": 1.092846393585205,
      "eval_runtime": 54.8964,
      "eval_samples_per_second": 91.263,
      "eval_steps_per_second": 11.422,
      "step": 11000
    },
    {
      "epoch": 0.49,
      "learning_rate": 3.3319444444444444e-07,
      "loss": 1.1024,
      "step": 12000
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.39261477045908183,
      "eval_loss": 1.089038372039795,
      "eval_runtime": 54.9763,
      "eval_samples_per_second": 91.13,
      "eval_steps_per_second": 11.405,
      "step": 12000
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.6094444444444446e-07,
      "loss": 1.1007,
      "step": 13000
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 0.34311377245508984,
      "eval_loss": 1.0933948755264282,
      "eval_runtime": 54.9285,
      "eval_samples_per_second": 91.209,
      "eval_steps_per_second": 11.415,
      "step": 13000
    },
    {
      "epoch": 0.57,
      "learning_rate": 3.8872222222222223e-07,
      "loss": 1.0985,
      "step": 14000
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.36367265469061877,
      "eval_loss": 1.09434974193573,
      "eval_runtime": 54.8032,
      "eval_samples_per_second": 91.418,
      "eval_steps_per_second": 11.441,
      "step": 14000
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.1650000000000006e-07,
      "loss": 1.0988,
      "step": 15000
    },
    {
      "epoch": 0.61,
      "eval_accuracy": 0.39481037924151696,
      "eval_loss": 1.0886671543121338,
      "eval_runtime": 54.9221,
      "eval_samples_per_second": 91.22,
      "eval_steps_per_second": 11.416,
      "step": 15000
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.4425e-07,
      "loss": 1.0965,
      "step": 16000
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.3916167664670659,
      "eval_loss": 1.0834949016571045,
      "eval_runtime": 54.5628,
      "eval_samples_per_second": 91.821,
      "eval_steps_per_second": 11.491,
      "step": 16000
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.7202777777777785e-07,
      "loss": 1.0926,
      "step": 17000
    },
    {
      "epoch": 0.69,
      "eval_accuracy": 0.4239520958083832,
      "eval_loss": 1.079688310623169,
      "eval_runtime": 54.6989,
      "eval_samples_per_second": 91.592,
      "eval_steps_per_second": 11.463,
      "step": 17000
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.998055555555556e-07,
      "loss": 1.0956,
      "step": 18000
    },
    {
      "epoch": 0.73,
      "eval_accuracy": 0.4219560878243513,
      "eval_loss": 1.080493688583374,
      "eval_runtime": 54.6863,
      "eval_samples_per_second": 91.613,
      "eval_steps_per_second": 11.465,
      "step": 18000
    },
    {
      "epoch": 0.77,
      "learning_rate": 5.275555555555556e-07,
      "loss": 1.0878,
      "step": 19000
    },
    {
      "epoch": 0.77,
      "eval_accuracy": 0.4343313373253493,
      "eval_loss": 1.0664235353469849,
      "eval_runtime": 54.7843,
      "eval_samples_per_second": 91.45,
      "eval_steps_per_second": 11.445,
      "step": 19000
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.553333333333334e-07,
      "loss": 1.0793,
      "step": 20000
    },
    {
      "epoch": 0.81,
      "eval_accuracy": 0.4365269461077844,
      "eval_loss": 1.06390380859375,
      "eval_runtime": 54.7978,
      "eval_samples_per_second": 91.427,
      "eval_steps_per_second": 11.442,
      "step": 20000
    },
    {
      "epoch": 0.86,
      "learning_rate": 5.830833333333334e-07,
      "loss": 1.0746,
      "step": 21000
    },
    {
      "epoch": 0.86,
      "eval_accuracy": 0.4311377245508982,
      "eval_loss": 1.0611063241958618,
      "eval_runtime": 54.6084,
      "eval_samples_per_second": 91.744,
      "eval_steps_per_second": 11.482,
      "step": 21000
    },
    {
      "epoch": 0.9,
      "learning_rate": 6.108611111111111e-07,
      "loss": 1.0757,
      "step": 22000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.43253493013972055,
      "eval_loss": 1.0579031705856323,
      "eval_runtime": 54.7147,
      "eval_samples_per_second": 91.566,
      "eval_steps_per_second": 11.459,
      "step": 22000
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.386111111111112e-07,
      "loss": 1.0712,
      "step": 23000
    },
    {
      "epoch": 0.94,
      "eval_accuracy": 0.43353293413173655,
      "eval_loss": 1.0545520782470703,
      "eval_runtime": 54.8205,
      "eval_samples_per_second": 91.389,
      "eval_steps_per_second": 11.437,
      "step": 23000
    },
    {
      "epoch": 0.98,
      "learning_rate": 6.663888888888889e-07,
      "loss": 1.0703,
      "step": 24000
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.43872255489021955,
      "eval_loss": 1.0489881038665771,
      "eval_runtime": 54.6899,
      "eval_samples_per_second": 91.607,
      "eval_steps_per_second": 11.465,
      "step": 24000
    },
    {
      "epoch": 1.02,
      "learning_rate": 6.941666666666667e-07,
      "loss": 1.0673,
      "step": 25000
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.4467065868263473,
      "eval_loss": 1.0496938228607178,
      "eval_runtime": 54.8292,
      "eval_samples_per_second": 91.375,
      "eval_steps_per_second": 11.436,
      "step": 25000
    },
    {
      "epoch": 1.06,
      "learning_rate": 7.219166666666666e-07,
      "loss": 1.0672,
      "step": 26000
    },
    {
      "epoch": 1.06,
      "eval_accuracy": 0.4297405189620758,
      "eval_loss": 1.0511012077331543,
      "eval_runtime": 54.8046,
      "eval_samples_per_second": 91.416,
      "eval_steps_per_second": 11.441,
      "step": 26000
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.496944444444444e-07,
      "loss": 1.0658,
      "step": 27000
    },
    {
      "epoch": 1.1,
      "eval_accuracy": 0.43912175648702595,
      "eval_loss": 1.0465787649154663,
      "eval_runtime": 54.7507,
      "eval_samples_per_second": 91.506,
      "eval_steps_per_second": 11.452,
      "step": 27000
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.774444444444445e-07,
      "loss": 1.0638,
      "step": 28000
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.43333333333333335,
      "eval_loss": 1.0430774688720703,
      "eval_runtime": 54.6439,
      "eval_samples_per_second": 91.685,
      "eval_steps_per_second": 11.474,
      "step": 28000
    },
    {
      "epoch": 1.18,
      "learning_rate": 8.052222222222223e-07,
      "loss": 1.0602,
      "step": 29000
    },
    {
      "epoch": 1.18,
      "eval_accuracy": 0.4479041916167665,
      "eval_loss": 1.0387687683105469,
      "eval_runtime": 54.8009,
      "eval_samples_per_second": 91.422,
      "eval_steps_per_second": 11.441,
      "step": 29000
    },
    {
      "epoch": 1.22,
      "learning_rate": 8.329722222222223e-07,
      "loss": 1.0567,
      "step": 30000
    },
    {
      "epoch": 1.22,
      "eval_accuracy": 0.4485029940119761,
      "eval_loss": 1.0339411497116089,
      "eval_runtime": 54.6621,
      "eval_samples_per_second": 91.654,
      "eval_steps_per_second": 11.47,
      "step": 30000
    },
    {
      "epoch": 1.26,
      "learning_rate": 8.607500000000001e-07,
      "loss": 1.0611,
      "step": 31000
    },
    {
      "epoch": 1.26,
      "eval_accuracy": 0.4465069860279441,
      "eval_loss": 1.0385600328445435,
      "eval_runtime": 54.7106,
      "eval_samples_per_second": 91.573,
      "eval_steps_per_second": 11.46,
      "step": 31000
    },
    {
      "epoch": 1.3,
      "learning_rate": 8.885e-07,
      "loss": 1.0555,
      "step": 32000
    },
    {
      "epoch": 1.3,
      "eval_accuracy": 0.4377245508982036,
      "eval_loss": 1.0331332683563232,
      "eval_runtime": 54.7554,
      "eval_samples_per_second": 91.498,
      "eval_steps_per_second": 11.451,
      "step": 32000
    },
    {
      "epoch": 1.34,
      "learning_rate": 9.162777777777779e-07,
      "loss": 1.0512,
      "step": 33000
    },
    {
      "epoch": 1.34,
      "eval_accuracy": 0.44610778443113774,
      "eval_loss": 1.028577446937561,
      "eval_runtime": 54.7331,
      "eval_samples_per_second": 91.535,
      "eval_steps_per_second": 11.456,
      "step": 33000
    },
    {
      "epoch": 1.39,
      "learning_rate": 9.440277777777779e-07,
      "loss": 1.048,
      "step": 34000
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.4532934131736527,
      "eval_loss": 1.0286486148834229,
      "eval_runtime": 54.7585,
      "eval_samples_per_second": 91.493,
      "eval_steps_per_second": 11.45,
      "step": 34000
    },
    {
      "epoch": 1.43,
      "learning_rate": 9.718055555555557e-07,
      "loss": 1.0524,
      "step": 35000
    },
    {
      "epoch": 1.43,
      "eval_accuracy": 0.449500998003992,
      "eval_loss": 1.0262507200241089,
      "eval_runtime": 54.1381,
      "eval_samples_per_second": 92.541,
      "eval_steps_per_second": 11.582,
      "step": 35000
    },
    {
      "epoch": 1.47,
      "learning_rate": 9.995555555555557e-07,
      "loss": 1.0472,
      "step": 36000
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.4552894211576846,
      "eval_loss": 1.022876501083374,
      "eval_runtime": 54.2077,
      "eval_samples_per_second": 92.422,
      "eval_steps_per_second": 11.567,
      "step": 36000
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.0273333333333335e-06,
      "loss": 1.0454,
      "step": 37000
    },
    {
      "epoch": 1.51,
      "eval_accuracy": 0.4477045908183633,
      "eval_loss": 1.0219467878341675,
      "eval_runtime": 54.4001,
      "eval_samples_per_second": 92.095,
      "eval_steps_per_second": 11.526,
      "step": 37000
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.0550833333333334e-06,
      "loss": 1.0473,
      "step": 38000
    },
    {
      "epoch": 1.55,
      "eval_accuracy": 0.46107784431137727,
      "eval_loss": 1.0189749002456665,
      "eval_runtime": 54.291,
      "eval_samples_per_second": 92.28,
      "eval_steps_per_second": 11.549,
      "step": 38000
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.0828611111111111e-06,
      "loss": 1.0465,
      "step": 39000
    },
    {
      "epoch": 1.59,
      "eval_accuracy": 0.4500998003992016,
      "eval_loss": 1.0226292610168457,
      "eval_runtime": 54.6634,
      "eval_samples_per_second": 91.652,
      "eval_steps_per_second": 11.47,
      "step": 39000
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.1106111111111112e-06,
      "loss": 1.0408,
      "step": 40000
    },
    {
      "epoch": 1.63,
      "eval_accuracy": 0.45229540918163674,
      "eval_loss": 1.0191301107406616,
      "eval_runtime": 54.6371,
      "eval_samples_per_second": 91.696,
      "eval_steps_per_second": 11.476,
      "step": 40000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 10000000,
  "num_train_epochs": 408,
  "save_steps": 1000,
  "total_flos": 8.36139439203287e+16,
  "trial_name": null,
  "trial_params": null
}