{
  "best_metric": 3.599998950958252,
  "best_model_checkpoint": "models/GPT2_five_function_42/checkpoint-64080",
  "epoch": 10.0,
  "global_step": 64080,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "learning_rate": 1e-05,
      "loss": 6.9949,
      "step": 1000
    },
    {
      "epoch": 0.31,
      "learning_rate": 2e-05,
      "loss": 5.7461,
      "step": 2000
    },
    {
      "epoch": 0.47,
      "learning_rate": 3e-05,
      "loss": 5.4085,
      "step": 3000
    },
    {
      "epoch": 0.62,
      "learning_rate": 4e-05,
      "loss": 5.1434,
      "step": 4000
    },
    {
      "epoch": 0.78,
      "learning_rate": 5e-05,
      "loss": 4.933,
      "step": 5000
    },
    {
      "epoch": 0.94,
      "learning_rate": 6e-05,
      "loss": 4.7568,
      "step": 6000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3263572316618317,
      "eval_loss": 4.562152862548828,
      "eval_runtime": 1.9609,
      "eval_samples_per_second": 600.758,
      "eval_steps_per_second": 5.1,
      "step": 6408
    },
    {
      "epoch": 1.09,
      "learning_rate": 7e-05,
      "loss": 4.6067,
      "step": 7000
    },
    {
      "epoch": 1.25,
      "learning_rate": 8e-05,
      "loss": 4.4864,
      "step": 8000
    },
    {
      "epoch": 1.4,
      "learning_rate": 9e-05,
      "loss": 4.3914,
      "step": 9000
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0001,
      "loss": 4.3109,
      "step": 10000
    },
    {
      "epoch": 1.72,
      "learning_rate": 9.889000000000001e-05,
      "loss": 4.2433,
      "step": 11000
    },
    {
      "epoch": 1.87,
      "learning_rate": 9.77788888888889e-05,
      "loss": 4.1832,
      "step": 12000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.36074087937649557,
      "eval_loss": 4.056756019592285,
      "eval_runtime": 1.9676,
      "eval_samples_per_second": 598.695,
      "eval_steps_per_second": 5.082,
      "step": 12816
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.66688888888889e-05,
      "loss": 4.1265,
      "step": 13000
    },
    {
      "epoch": 2.18,
      "learning_rate": 9.55588888888889e-05,
      "loss": 4.0565,
      "step": 14000
    },
    {
      "epoch": 2.34,
      "learning_rate": 9.444777777777778e-05,
      "loss": 4.0283,
      "step": 15000
    },
    {
      "epoch": 2.5,
      "learning_rate": 9.333777777777777e-05,
      "loss": 4.0035,
      "step": 16000
    },
    {
      "epoch": 2.65,
      "learning_rate": 9.222666666666668e-05,
      "loss": 3.9768,
      "step": 17000
    },
    {
      "epoch": 2.81,
      "learning_rate": 9.111666666666667e-05,
      "loss": 3.9553,
      "step": 18000
    },
    {
      "epoch": 2.97,
      "learning_rate": 9.000555555555557e-05,
      "loss": 3.9365,
      "step": 19000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3769100169779287,
      "eval_loss": 3.874519109725952,
      "eval_runtime": 1.9728,
      "eval_samples_per_second": 597.128,
      "eval_steps_per_second": 5.069,
      "step": 19224
    },
    {
      "epoch": 3.12,
      "learning_rate": 8.889555555555556e-05,
      "loss": 3.871,
      "step": 20000
    },
    {
      "epoch": 3.28,
      "learning_rate": 8.778444444444445e-05,
      "loss": 3.8532,
      "step": 21000
    },
    {
      "epoch": 3.43,
      "learning_rate": 8.667444444444444e-05,
      "loss": 3.8443,
      "step": 22000
    },
    {
      "epoch": 3.59,
      "learning_rate": 8.556444444444445e-05,
      "loss": 3.8378,
      "step": 23000
    },
    {
      "epoch": 3.75,
      "learning_rate": 8.445333333333333e-05,
      "loss": 3.825,
      "step": 24000
    },
    {
      "epoch": 3.9,
      "learning_rate": 8.334222222222222e-05,
      "loss": 3.8154,
      "step": 25000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3853922971003837,
      "eval_loss": 3.779280185699463,
      "eval_runtime": 1.9858,
      "eval_samples_per_second": 593.214,
      "eval_steps_per_second": 5.036,
      "step": 25632
    },
    {
      "epoch": 4.06,
      "learning_rate": 8.223111111111111e-05,
      "loss": 3.7779,
      "step": 26000
    },
    {
      "epoch": 4.21,
      "learning_rate": 8.112111111111111e-05,
      "loss": 3.7386,
      "step": 27000
    },
    {
      "epoch": 4.37,
      "learning_rate": 8.001e-05,
      "loss": 3.7408,
      "step": 28000
    },
    {
      "epoch": 4.53,
      "learning_rate": 7.890000000000001e-05,
      "loss": 3.7369,
      "step": 29000
    },
    {
      "epoch": 4.68,
      "learning_rate": 7.77888888888889e-05,
      "loss": 3.7354,
      "step": 30000
    },
    {
      "epoch": 4.84,
      "learning_rate": 7.667888888888889e-05,
      "loss": 3.7297,
      "step": 31000
    },
    {
      "epoch": 4.99,
      "learning_rate": 7.556777777777779e-05,
      "loss": 3.725,
      "step": 32000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3909669398286165,
      "eval_loss": 3.7228105068206787,
      "eval_runtime": 2.0428,
      "eval_samples_per_second": 576.664,
      "eval_steps_per_second": 4.895,
      "step": 32040
    },
    {
      "epoch": 5.15,
      "learning_rate": 7.445777777777778e-05,
      "loss": 3.6547,
      "step": 33000
    },
    {
      "epoch": 5.31,
      "learning_rate": 7.334666666666668e-05,
      "loss": 3.6602,
      "step": 34000
    },
    {
      "epoch": 5.46,
      "learning_rate": 7.223666666666667e-05,
      "loss": 3.6624,
      "step": 35000
    },
    {
      "epoch": 5.62,
      "learning_rate": 7.112555555555556e-05,
      "loss": 3.6649,
      "step": 36000
    },
    {
      "epoch": 5.77,
      "learning_rate": 7.001555555555555e-05,
      "loss": 3.6615,
      "step": 37000
    },
    {
      "epoch": 5.93,
      "learning_rate": 6.890444444444445e-05,
      "loss": 3.6621,
      "step": 38000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.39523147467347564,
      "eval_loss": 3.681443691253662,
      "eval_runtime": 1.9586,
      "eval_samples_per_second": 601.465,
      "eval_steps_per_second": 5.106,
      "step": 38448
    },
    {
      "epoch": 6.09,
      "learning_rate": 6.779444444444444e-05,
      "loss": 3.6176,
      "step": 39000
    },
    {
      "epoch": 6.24,
      "learning_rate": 6.668333333333333e-05,
      "loss": 3.5948,
      "step": 40000
    },
    {
      "epoch": 6.4,
      "learning_rate": 6.557333333333332e-05,
      "loss": 3.6021,
      "step": 41000
    },
    {
      "epoch": 6.55,
      "learning_rate": 6.446222222222223e-05,
      "loss": 3.6043,
      "step": 42000
    },
    {
      "epoch": 6.71,
      "learning_rate": 6.335222222222222e-05,
      "loss": 3.6064,
      "step": 43000
    },
    {
      "epoch": 6.87,
      "learning_rate": 6.224111111111112e-05,
      "loss": 3.6072,
      "step": 44000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.39868721842706845,
      "eval_loss": 3.6536548137664795,
      "eval_runtime": 1.9678,
      "eval_samples_per_second": 598.633,
      "eval_steps_per_second": 5.082,
      "step": 44856
    },
    {
      "epoch": 7.02,
      "learning_rate": 6.113111111111111e-05,
      "loss": 3.5958,
      "step": 45000
    },
    {
      "epoch": 7.18,
      "learning_rate": 6.002e-05,
      "loss": 3.5371,
      "step": 46000
    },
    {
      "epoch": 7.33,
      "learning_rate": 5.891e-05,
      "loss": 3.5493,
      "step": 47000
    },
    {
      "epoch": 7.49,
      "learning_rate": 5.779888888888889e-05,
      "loss": 3.554,
      "step": 48000
    },
    {
      "epoch": 7.65,
      "learning_rate": 5.668888888888889e-05,
      "loss": 3.5593,
      "step": 49000
    },
    {
      "epoch": 7.8,
      "learning_rate": 5.5577777777777784e-05,
      "loss": 3.5597,
      "step": 50000
    },
    {
      "epoch": 7.96,
      "learning_rate": 5.446777777777778e-05,
      "loss": 3.5598,
      "step": 51000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.4002713794901274,
      "eval_loss": 3.631843328475952,
      "eval_runtime": 1.9682,
      "eval_samples_per_second": 598.504,
      "eval_steps_per_second": 5.081,
      "step": 51264
    },
    {
      "epoch": 8.11,
      "learning_rate": 5.3356666666666663e-05,
      "loss": 3.5064,
      "step": 52000
    },
    {
      "epoch": 8.27,
      "learning_rate": 5.224666666666667e-05,
      "loss": 3.5007,
      "step": 53000
    },
    {
      "epoch": 8.43,
      "learning_rate": 5.1135555555555556e-05,
      "loss": 3.511,
      "step": 54000
    },
    {
      "epoch": 8.58,
      "learning_rate": 5.002555555555556e-05,
      "loss": 3.5135,
      "step": 55000
    },
    {
      "epoch": 8.74,
      "learning_rate": 4.891444444444444e-05,
      "loss": 3.5193,
      "step": 56000
    },
    {
      "epoch": 8.9,
      "learning_rate": 4.780444444444445e-05,
      "loss": 3.5203,
      "step": 57000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.4027378581072952,
      "eval_loss": 3.6141819953918457,
      "eval_runtime": 1.9786,
      "eval_samples_per_second": 595.365,
      "eval_steps_per_second": 5.054,
      "step": 57672
    },
    {
      "epoch": 9.05,
      "learning_rate": 4.6693333333333336e-05,
      "loss": 3.4957,
      "step": 58000
    },
    {
      "epoch": 9.21,
      "learning_rate": 4.5583333333333335e-05,
      "loss": 3.4579,
      "step": 59000
    },
    {
      "epoch": 9.36,
      "learning_rate": 4.447222222222223e-05,
      "loss": 3.4682,
      "step": 60000
    },
    {
      "epoch": 9.52,
      "learning_rate": 4.336222222222223e-05,
      "loss": 3.4743,
      "step": 61000
    },
    {
      "epoch": 9.68,
      "learning_rate": 4.2251111111111115e-05,
      "loss": 3.4784,
      "step": 62000
    },
    {
      "epoch": 9.83,
      "learning_rate": 4.1141111111111114e-05,
      "loss": 3.4833,
      "step": 63000
    },
    {
      "epoch": 9.99,
      "learning_rate": 4.003e-05,
      "loss": 3.4839,
      "step": 64000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.40503723112709383,
      "eval_loss": 3.599998950958252,
      "eval_runtime": 1.9622,
      "eval_samples_per_second": 600.357,
      "eval_steps_per_second": 5.096,
      "step": 64080
    }
  ],
  "max_steps": 100000,
  "num_train_epochs": 16,
  "total_flos": 5.3573097259008e+17,
  "trial_name": null,
  "trial_params": null
}