{
  "best_metric": 0.9126,
  "best_model_checkpoint": "../../checkpoint/imdb/roberta-large/checkpoint-14858",
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 15640,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.89904,
      "eval_loss": 0.2406211644411087,
      "eval_runtime": 103.5694,
      "eval_samples_per_second": 241.384,
      "eval_steps_per_second": 0.946,
      "step": 782
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.680306905370844e-05,
      "loss": 0.3075,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8962,
      "eval_loss": 0.2538779377937317,
      "eval_runtime": 102.8426,
      "eval_samples_per_second": 243.09,
      "eval_steps_per_second": 0.953,
      "step": 1564
    },
    {
      "epoch": 2.56,
      "learning_rate": 4.360613810741688e-05,
      "loss": 0.2171,
      "step": 2000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.90308,
      "eval_loss": 0.2649693489074707,
      "eval_runtime": 102.9465,
      "eval_samples_per_second": 242.844,
      "eval_steps_per_second": 0.952,
      "step": 2346
    },
    {
      "epoch": 3.84,
      "learning_rate": 4.040920716112532e-05,
      "loss": 0.1697,
      "step": 3000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.89728,
      "eval_loss": 0.3427022695541382,
      "eval_runtime": 103.0024,
      "eval_samples_per_second": 242.713,
      "eval_steps_per_second": 0.951,
      "step": 3128
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.90308,
      "eval_loss": 0.3241328001022339,
      "eval_runtime": 102.9184,
      "eval_samples_per_second": 242.911,
      "eval_steps_per_second": 0.952,
      "step": 3910
    },
    {
      "epoch": 5.12,
      "learning_rate": 3.721227621483376e-05,
      "loss": 0.1339,
      "step": 4000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.90492,
      "eval_loss": 0.41408097743988037,
      "eval_runtime": 102.8464,
      "eval_samples_per_second": 243.081,
      "eval_steps_per_second": 0.953,
      "step": 4692
    },
    {
      "epoch": 6.39,
      "learning_rate": 3.40153452685422e-05,
      "loss": 0.1038,
      "step": 5000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8946,
      "eval_loss": 0.45717746019363403,
      "eval_runtime": 102.3334,
      "eval_samples_per_second": 244.299,
      "eval_steps_per_second": 0.958,
      "step": 5474
    },
    {
      "epoch": 7.67,
      "learning_rate": 3.081841432225064e-05,
      "loss": 0.0922,
      "step": 6000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9054,
      "eval_loss": 0.4153657555580139,
      "eval_runtime": 102.2979,
      "eval_samples_per_second": 244.384,
      "eval_steps_per_second": 0.958,
      "step": 6256
    },
    {
      "epoch": 8.95,
      "learning_rate": 2.7621483375959077e-05,
      "loss": 0.0676,
      "step": 7000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.89824,
      "eval_loss": 0.5019603371620178,
      "eval_runtime": 102.1481,
      "eval_samples_per_second": 244.743,
      "eval_steps_per_second": 0.959,
      "step": 7038
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.90708,
      "eval_loss": 0.507008969783783,
      "eval_runtime": 102.4086,
      "eval_samples_per_second": 244.12,
      "eval_steps_per_second": 0.957,
      "step": 7820
    },
    {
      "epoch": 10.23,
      "learning_rate": 2.442455242966752e-05,
      "loss": 0.0568,
      "step": 8000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.90668,
      "eval_loss": 0.48255667090415955,
      "eval_runtime": 102.2755,
      "eval_samples_per_second": 244.438,
      "eval_steps_per_second": 0.958,
      "step": 8602
    },
    {
      "epoch": 11.51,
      "learning_rate": 2.122762148337596e-05,
      "loss": 0.0443,
      "step": 9000
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.90856,
      "eval_loss": 0.5103762745857239,
      "eval_runtime": 102.4293,
      "eval_samples_per_second": 244.071,
      "eval_steps_per_second": 0.957,
      "step": 9384
    },
    {
      "epoch": 12.79,
      "learning_rate": 1.80306905370844e-05,
      "loss": 0.0313,
      "step": 10000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.90884,
      "eval_loss": 0.545563817024231,
      "eval_runtime": 101.9339,
      "eval_samples_per_second": 245.257,
      "eval_steps_per_second": 0.961,
      "step": 10166
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.90776,
      "eval_loss": 0.47395065426826477,
      "eval_runtime": 102.2186,
      "eval_samples_per_second": 244.574,
      "eval_steps_per_second": 0.959,
      "step": 10948
    },
    {
      "epoch": 14.07,
      "learning_rate": 1.483375959079284e-05,
      "loss": 0.0245,
      "step": 11000
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.90712,
      "eval_loss": 0.4977429211139679,
      "eval_runtime": 102.2005,
      "eval_samples_per_second": 244.617,
      "eval_steps_per_second": 0.959,
      "step": 11730
    },
    {
      "epoch": 15.35,
      "learning_rate": 1.163682864450128e-05,
      "loss": 0.0227,
      "step": 12000
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.90984,
      "eval_loss": 0.5135776400566101,
      "eval_runtime": 102.5324,
      "eval_samples_per_second": 243.825,
      "eval_steps_per_second": 0.956,
      "step": 12512
    },
    {
      "epoch": 16.62,
      "learning_rate": 8.439897698209718e-06,
      "loss": 0.0175,
      "step": 13000
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.91076,
      "eval_loss": 0.5130705833435059,
      "eval_runtime": 102.0766,
      "eval_samples_per_second": 244.914,
      "eval_steps_per_second": 0.96,
      "step": 13294
    },
    {
      "epoch": 17.9,
      "learning_rate": 5.242966751918159e-06,
      "loss": 0.0173,
      "step": 14000
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.91092,
      "eval_loss": 0.5370119214057922,
      "eval_runtime": 102.316,
      "eval_samples_per_second": 244.341,
      "eval_steps_per_second": 0.958,
      "step": 14076
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.9126,
      "eval_loss": 0.5343945026397705,
      "eval_runtime": 102.4054,
      "eval_samples_per_second": 244.128,
      "eval_steps_per_second": 0.957,
      "step": 14858
    },
    {
      "epoch": 19.18,
      "learning_rate": 2.0460358056265987e-06,
      "loss": 0.0152,
      "step": 15000
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.91212,
      "eval_loss": 0.5329343676567078,
      "eval_runtime": 102.1338,
      "eval_samples_per_second": 244.777,
      "eval_steps_per_second": 0.96,
      "step": 15640
    },
    {
      "epoch": 20.0,
      "step": 15640,
      "total_flos": 1.16491420416e+17,
      "train_loss": 0.08507246678442601,
      "train_runtime": 7726.1299,
      "train_samples_per_second": 64.715,
      "train_steps_per_second": 2.024
    }
  ],
  "logging_steps": 1000,
  "max_steps": 15640,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 1.16491420416e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}