trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0029940119760479044,
  "global_step": 1,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 5e-05,
      "loss": 41.4703,
      "step": 1
    },
    {
      "epoch": 0.0,
      "eval_loss": 44.2744026184082,
      "eval_runtime": 2.632,
      "eval_samples_per_second": 166.03,
      "eval_steps_per_second": 14.057,
      "step": 1
    },
    {
      "epoch": 0.0,
      "step": 1,
      "total_flos": 3135561007104.0,
      "train_loss": 41.470333099365234,
      "train_runtime": 8.8744,
      "train_samples_per_second": 1.352,
      "train_steps_per_second": 0.113
    }
  ],
  "max_steps": 1,
  "num_train_epochs": 1,
  "total_flos": 3135561007104.0,
  "trial_name": null,
  "trial_params": null
}