{
  "best_metric": 2.9109926223754883,
  "best_model_checkpoint": "miner_id_24/checkpoint-10",
  "epoch": 0.05486968449931413,
  "eval_steps": 5,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0054869684499314125,
      "eval_loss": 2.944849967956543,
      "eval_runtime": 3.8621,
      "eval_samples_per_second": 39.874,
      "eval_steps_per_second": 10.098,
      "step": 1
    },
    {
      "epoch": 0.01646090534979424,
      "grad_norm": 27.84040641784668,
      "learning_rate": 0.00012,
      "loss": 11.4365,
      "step": 3
    },
    {
      "epoch": 0.027434842249657063,
      "eval_loss": 2.921481132507324,
      "eval_runtime": 2.8019,
      "eval_samples_per_second": 54.963,
      "eval_steps_per_second": 13.919,
      "step": 5
    },
    {
      "epoch": 0.03292181069958848,
      "grad_norm": 22.27212905883789,
      "learning_rate": 0.0001992114701314478,
      "loss": 11.4997,
      "step": 6
    },
    {
      "epoch": 0.04938271604938271,
      "grad_norm": 26.73714256286621,
      "learning_rate": 0.00018763066800438636,
      "loss": 11.8003,
      "step": 9
    },
    {
      "epoch": 0.05486968449931413,
      "eval_loss": 2.9109926223754883,
      "eval_runtime": 2.6688,
      "eval_samples_per_second": 57.703,
      "eval_steps_per_second": 14.613,
      "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": 1,
        "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": 415957801500672.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}