checkpoint-131/trainer_state.json

{
  "best_metric": 0.21770058572292328,
  "best_model_checkpoint": "imdb_full/checkpoint-131",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 131,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9204,
      "eval_loss": 0.21770058572292328,
      "eval_runtime": 78.2774,
      "eval_samples_per_second": 319.377,
      "eval_steps_per_second": 1.674,
      "step": 131
    }
  ],
  "logging_steps": 500,
  "max_steps": 655,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1.30648375296e+16,
  "train_batch_size": 192,
  "trial_name": null,
  "trial_params": null
}