trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4848719937936385,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.48,
      "learning_rate": 2.576852211016292e-05,
      "loss": 2.3941,
      "step": 10000
    },
    {
      "epoch": 0.48,
      "eval_accuracy": 0.46707198326043414,
      "eval_loss": 2.4748170375823975,
      "eval_runtime": 895.2093,
      "eval_samples_per_second": 11.522,
      "eval_steps_per_second": 5.762,
      "step": 10000
    },
    {
      "epoch": 0.48,
      "step": 10000,
      "total_flos": 3.093389377536e+17,
      "train_loss": 0.0,
      "train_runtime": 139.4935,
      "train_samples_per_second": 295.691,
      "train_steps_per_second": 4.617
    }
  ],
  "max_steps": 644,
  "num_train_epochs": 1,
  "total_flos": 3.093389377536e+17,
  "trial_name": null,
  "trial_params": null
}