trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8571428571428571,
  "eval_steps": 500,
  "global_step": 3,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 7.499046883341479,
      "learning_rate": 2e-05,
      "loss": 0.862,
      "mean_token_accuracy": 0.7828678265213966,
      "num_tokens": 524288.0,
      "step": 1
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 4.840174544742125,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7242,
      "mean_token_accuracy": 0.7936247298493981,
      "num_tokens": 1045346.0,
      "step": 2
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 130.3488135782814,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.7091,
      "mean_token_accuracy": 0.8003682056441903,
      "num_tokens": 1569634.0,
      "step": 3
    },
    {
      "epoch": 0.8571428571428571,
      "step": 3,
      "total_flos": 3122684100608.0,
      "train_loss": 0.7650859951972961,
      "train_runtime": 556.5652,
      "train_samples_per_second": 0.805,
      "train_steps_per_second": 0.005
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 3,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3122684100608.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}