trainer_state.json

{
  "best_metric": 2.3684260845184326,
  "best_model_checkpoint": "models/checkpoint-303",
  "epoch": 2.9914110429447853,
  "global_step": 303,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.99,
      "eval_accuracy": 0.48085171490501083,
      "eval_loss": 2.563568353652954,
      "eval_runtime": 9.7149,
      "eval_samples_per_second": 59.187,
      "eval_steps_per_second": 7.411,
      "step": 101
    },
    {
      "epoch": 1.99,
      "eval_accuracy": 0.501789281312423,
      "eval_loss": 2.4074697494506836,
      "eval_runtime": 9.6785,
      "eval_samples_per_second": 59.41,
      "eval_steps_per_second": 7.439,
      "step": 202
    },
    {
      "epoch": 2.99,
      "eval_accuracy": 0.5070321730630286,
      "eval_loss": 2.3684260845184326,
      "eval_runtime": 9.6634,
      "eval_samples_per_second": 59.503,
      "eval_steps_per_second": 7.451,
      "step": 303
    },
    {
      "epoch": 2.99,
      "step": 303,
      "total_flos": 5095194624000000.0,
      "train_loss": 2.700293814781869,
      "train_runtime": 444.9433,
      "train_samples_per_second": 21.974,
      "train_steps_per_second": 0.681
    }
  ],
  "max_steps": 303,
  "num_train_epochs": 3,
  "total_flos": 5095194624000000.0,
  "trial_name": null,
  "trial_params": null
}