{
  "best_metric": 1.335775375366211,
  "best_model_checkpoint": "miner_id_24/checkpoint-10",
  "epoch": 0.00016986364196141547,
  "eval_steps": 5,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.6986364196141547e-05,
      "eval_loss": 1.3468016386032104,
      "eval_runtime": 3825.0823,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 3.24,
      "step": 1
    },
    {
      "epoch": 5.095909258842464e-05,
      "grad_norm": 15.381132125854492,
      "learning_rate": 3e-05,
      "loss": 5.8377,
      "step": 3
    },
    {
      "epoch": 8.493182098070774e-05,
      "eval_loss": 1.34633207321167,
      "eval_runtime": 3844.7283,
      "eval_samples_per_second": 6.447,
      "eval_steps_per_second": 3.224,
      "step": 5
    },
    {
      "epoch": 0.00010191818517684929,
      "grad_norm": 13.268209457397461,
      "learning_rate": 6e-05,
      "loss": 5.5592,
      "step": 6
    },
    {
      "epoch": 0.00015287727776527394,
      "grad_norm": 38.36406707763672,
      "learning_rate": 9e-05,
      "loss": 5.5275,
      "step": 9
    },
    {
      "epoch": 0.00016986364196141547,
      "eval_loss": 1.335775375366211,
      "eval_runtime": 3844.3305,
      "eval_samples_per_second": 6.448,
      "eval_steps_per_second": 3.224,
      "step": 10
    }
  ],
  "logging_steps": 3,
  "max_steps": 30,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 2,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3349006302314496.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}