checkpoint-50/trainer_state.json

{
  "best_metric": NaN,
  "best_model_checkpoint": "outputllama7-4/checkpoint-50",
  "epoch": 0.4276114844227245,
  "eval_steps": 50,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.43,
      "learning_rate": 0.0,
      "loss": 0.0,
      "step": 50
    },
    {
      "epoch": 0.43,
      "eval_loss": NaN,
      "eval_runtime": 27.7284,
      "eval_samples_per_second": 29.537,
      "eval_steps_per_second": 9.846,
      "step": 50
    }
  ],
  "logging_steps": 50,
  "max_steps": 350,
  "num_train_epochs": 4,
  "save_steps": 50,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}