checkpoint-6/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 6,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "learning_rate": 3e-05,
      "loss": 2.4435,
      "step": 1
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.4e-05,
      "loss": 2.3269,
      "step": 2
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.8e-05,
      "loss": 2.2858,
      "step": 3
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.2e-05,
      "loss": 2.3555,
      "step": 4
    },
    {
      "epoch": 2.5,
      "learning_rate": 6e-06,
      "loss": 2.2663,
      "step": 5
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0,
      "loss": 2.2476,
      "step": 6
    }
  ],
  "logging_steps": 1,
  "max_steps": 6,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 487764705411072.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}