trainer_state.json

{
  "best_metric": 0.9143889652364229,
  "best_model_checkpoint": "./resultsv3p2/output/checkpoint-196352",
  "epoch": 2.0,
  "global_step": 196352,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 3.1928048978707266e-06,
      "loss": 0.4695,
      "step": 98176
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9100117066218761,
      "eval_loss": 0.3915119469165802,
      "eval_runtime": 352.6229,
      "eval_samples_per_second": 55.717,
      "eval_steps_per_second": 13.93,
      "step": 98176
    },
    {
      "epoch": 2.0,
      "learning_rate": 2.4055913745462424e-09,
      "loss": 0.2675,
      "step": 196352
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9143889652364229,
      "eval_loss": 0.4945842921733856,
      "eval_runtime": 227.1799,
      "eval_samples_per_second": 86.482,
      "eval_steps_per_second": 21.622,
      "step": 196352
    }
  ],
  "max_steps": 196352,
  "num_train_epochs": 2,
  "total_flos": 8.31341361422996e+16,
  "trial_name": null,
  "trial_params": null
}