{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 36.231884057971016,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 7.246376811594203,
      "grad_norm": 4.833008289337158,
      "learning_rate": 4.094202898550725e-05,
      "loss": 1.952,
      "step": 500
    },
    {
      "epoch": 14.492753623188406,
      "grad_norm": 4.0949931144714355,
      "learning_rate": 3.188405797101449e-05,
      "loss": 1.3781,
      "step": 1000
    },
    {
      "epoch": 21.73913043478261,
      "grad_norm": 2.9236912727355957,
      "learning_rate": 2.282608695652174e-05,
      "loss": 0.9801,
      "step": 1500
    },
    {
      "epoch": 28.985507246376812,
      "grad_norm": 2.521008014678955,
      "learning_rate": 1.3768115942028985e-05,
      "loss": 0.7325,
      "step": 2000
    },
    {
      "epoch": 36.231884057971016,
      "grad_norm": 2.6362154483795166,
      "learning_rate": 4.710144927536232e-06,
      "loss": 0.6039,
      "step": 2500
    }
  ],
  "logging_steps": 500,
  "max_steps": 2760,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 500,
  "total_flos": 2612920320000000.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}