{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.5,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "grad_norm": 13.729061126708984,
      "learning_rate": 8.336666666666668e-06,
      "loss": 0.222,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.73640251159668,
      "learning_rate": 6.6716666666666674e-06,
      "loss": 0.1257,
      "step": 2000
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.6222492456436157,
      "learning_rate": 5.0049999999999995e-06,
      "loss": 0.0851,
      "step": 3000
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.9226888418197632,
      "learning_rate": 3.3383333333333333e-06,
      "loss": 0.0732,
      "step": 4000
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.6427457928657532,
      "learning_rate": 1.6716666666666666e-06,
      "loss": 0.0744,
      "step": 5000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 6000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}