trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "global_step": 550,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 3e-05,
      "loss": 1.5605,
      "step": 275
    },
    {
      "epoch": 1.0,
      "eval_exact_match": 51.61478372391279,
      "eval_f1": 61.15406670627152,
      "eval_runtime": 8.3255,
      "eval_samples_per_second": 1030.202,
      "eval_steps_per_second": 3.243,
      "step": 275
    },
    {
      "epoch": 2.0,
      "learning_rate": 3e-05,
      "loss": 0.8136,
      "step": 550
    },
    {
      "epoch": 2.0,
      "eval_exact_match": 52.43091990206366,
      "eval_f1": 62.03808463298606,
      "eval_runtime": 7.4982,
      "eval_samples_per_second": 1143.88,
      "eval_steps_per_second": 3.601,
      "step": 550
    }
  ],
  "max_steps": 2750,
  "num_train_epochs": 10,
  "total_flos": 174079118147584.0,
  "trial_name": null,
  "trial_params": null
}