trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.01569242840329541,
  "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": 0.0002,
      "loss": 1.4673,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.2893,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.1835,
      "step": 30
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0317,
      "step": 40
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0207,
      "step": 50
    },
    {
      "epoch": 0.02,
      "step": 50,
      "total_flos": 2292376788271104.0,
      "train_loss": 1.198510036468506,
      "train_runtime": 688.023,
      "train_samples_per_second": 1.163,
      "train_steps_per_second": 0.073
    }
  ],
  "max_steps": 50,
  "num_train_epochs": 1,
  "total_flos": 2292376788271104.0,
  "trial_name": null,
  "trial_params": null
}