trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 324,
  "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": 1.1059,
      "step": 108
    },
    {
      "epoch": 1.0,
      "eval_exact_match": 71.69173370642416,
      "eval_f1": 75.22461569670578,
      "eval_runtime": 3.6341,
      "eval_samples_per_second": 2360.171,
      "eval_steps_per_second": 7.43,
      "step": 108
    },
    {
      "epoch": 2.0,
      "learning_rate": 3e-05,
      "loss": 0.4847,
      "step": 216
    },
    {
      "epoch": 2.0,
      "eval_exact_match": 72.01818817768451,
      "eval_f1": 75.8243729523036,
      "eval_runtime": 3.0173,
      "eval_samples_per_second": 2842.65,
      "eval_steps_per_second": 8.949,
      "step": 216
    },
    {
      "epoch": 3.0,
      "learning_rate": 3e-05,
      "loss": 0.4307,
      "step": 324
    },
    {
      "epoch": 3.0,
      "eval_exact_match": 73.68543779876414,
      "eval_f1": 76.7259721999141,
      "eval_runtime": 2.9802,
      "eval_samples_per_second": 2878.0,
      "eval_steps_per_second": 9.06,
      "step": 324
    }
  ],
  "max_steps": 2160,
  "num_train_epochs": 20,
  "total_flos": 38745933873152.0,
  "trial_name": null,
  "trial_params": null
}