trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.006980802792321117,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 0.016,
      "loss": 5.6438,
      "step": 10
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.012,
      "loss": 5.329,
      "step": 20
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.008,
      "loss": 5.1696,
      "step": 30
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.004,
      "loss": 5.0505,
      "step": 40
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0,
      "loss": 5.029,
      "step": 50
    }
  ],
  "max_steps": 50,
  "num_train_epochs": 1,
  "total_flos": 1733286061670400.0,
  "trial_name": null,
  "trial_params": null
}