checkpoint-18/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.769230769230769,
  "eval_steps": 500,
  "global_step": 18,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 4.8395,
      "step": 2
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.000175,
      "loss": 4.776,
      "step": 4
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00015000000000000001,
      "loss": 4.773,
      "step": 6
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.000125,
      "loss": 4.6848,
      "step": 8
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0001,
      "loss": 4.7228,
      "step": 10
    },
    {
      "epoch": 1.85,
      "learning_rate": 7.500000000000001e-05,
      "loss": 4.6165,
      "step": 12
    },
    {
      "epoch": 2.15,
      "learning_rate": 5e-05,
      "loss": 4.6521,
      "step": 14
    },
    {
      "epoch": 2.46,
      "learning_rate": 2.5e-05,
      "loss": 4.549,
      "step": 16
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.0,
      "loss": 4.5876,
      "step": 18
    }
  ],
  "logging_steps": 2,
  "max_steps": 18,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 1771363312336896.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}