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": 1.4911,
      "step": 275
    },
    {
      "epoch": 1.0,
      "eval_exact_match": 54.57619214177451,
      "eval_f1": 62.91955558112226,
      "eval_runtime": 8.6006,
      "eval_samples_per_second": 997.252,
      "eval_steps_per_second": 3.139,
      "step": 275
    },
    {
      "epoch": 2.0,
      "learning_rate": 3e-05,
      "loss": 0.7967,
      "step": 550
    },
    {
      "epoch": 2.0,
      "eval_exact_match": 57.619214177451326,
      "eval_f1": 64.82137371992548,
      "eval_runtime": 7.4409,
      "eval_samples_per_second": 1152.682,
      "eval_steps_per_second": 3.629,
      "step": 550
    }
  ],
  "max_steps": 2750,
  "num_train_epochs": 10,
  "total_flos": 174079118147584.0,
  "trial_name": null,
  "trial_params": null
}