trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.6611295681063123,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8305647840531561,
      "grad_norm": 0.341796875,
      "learning_rate": 5.555555555555556e-05,
      "loss": 2.1003,
      "step": 500
    },
    {
      "epoch": 1.6611295681063123,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0,
      "loss": 1.9472,
      "step": 1000
    }
  ],
  "logging_steps": 500,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 5.608559188821934e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}