last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.04955947136563876,
  "eval_steps": 10,
  "global_step": 90,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 4.2000000000000006e-07,
      "loss": 3.1165,
      "step": 25
    },
    {
      "epoch": 0.03,
      "learning_rate": 9.200000000000001e-07,
      "loss": 2.3577,
      "step": 50
    },
    {
      "epoch": 0.04,
      "learning_rate": 1.42e-06,
      "loss": 1.7546,
      "step": 75
    }
  ],
  "logging_steps": 25,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 10,
  "total_flos": 4.155629764608e+17,
  "trial_name": null,
  "trial_params": null
}