last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.08528085280852808,
  "eval_steps": 10,
  "global_step": 13,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006560065600656006,
      "eval_loss": NaN,
      "eval_runtime": 23.3316,
      "eval_samples_per_second": 44.018,
      "eval_steps_per_second": 5.529,
      "step": 1
    },
    {
      "epoch": 0.06560065600656007,
      "grad_norm": NaN,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.0,
      "step": 10
    },
    {
      "epoch": 0.06560065600656007,
      "eval_loss": NaN,
      "eval_runtime": 22.0132,
      "eval_samples_per_second": 46.654,
      "eval_steps_per_second": 5.86,
      "step": 10
    }
  ],
  "logging_steps": 10,
  "max_steps": 50,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 13,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3703987614253056.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}