trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0009174311926605505,
  "eval_steps": 500,
  "global_step": 1,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2e-05,
      "loss": 1.6972,
      "step": 1
    },
    {
      "epoch": 0.0,
      "eval_loss": 1.711571216583252,
      "eval_runtime": 0.5771,
      "eval_samples_per_second": 110.898,
      "eval_steps_per_second": 1.733,
      "step": 1
    },
    {
      "epoch": 0.0,
      "step": 1,
      "total_flos": 418759311360.0,
      "train_loss": 1.6971666812896729,
      "train_runtime": 12.3219,
      "train_samples_per_second": 10.388,
      "train_steps_per_second": 0.081
    }
  ],
  "logging_steps": 5,
  "max_steps": 1,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 418759311360.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}