{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 92,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.87,
      "grad_norm": 0.5561133027076721,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.3946,
      "step": 20
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.4310612976551056,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.3892,
      "step": 40
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.4552713632583618,
      "learning_rate": 6e-06,
      "loss": 0.3918,
      "step": 60
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.38019290566444397,
      "learning_rate": 7.9e-06,
      "loss": 0.345,
      "step": 80
    }
  ],
  "logging_steps": 20,
  "max_steps": 92,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 431142186516480.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}