trainer_state.json

{
  "best_metric": 4.054861068725586,
  "best_model_checkpoint": "output/john-k-samson/checkpoint-10",
  "epoch": 1.0,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "learning_rate": 6.86e-05,
      "loss": 4.4485,
      "step": 5
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0,
      "loss": 4.1856,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_loss": 4.054861068725586,
      "eval_runtime": 0.5716,
      "eval_samples_per_second": 22.743,
      "eval_steps_per_second": 3.499,
      "step": 10
    }
  ],
  "max_steps": 10,
  "num_train_epochs": 1,
  "total_flos": 10321035264000.0,
  "trial_name": null,
  "trial_params": null
}