{
  "best_metric": 0.39871343970298767,
  "best_model_checkpoint": "mi-banana-variedades/checkpoint-16",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 16,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "grad_norm": 11.00576114654541,
      "learning_rate": 2.5e-05,
      "loss": 0.5206,
      "step": 1
    },
    {
      "epoch": 1.0,
      "grad_norm": 24.502017974853516,
      "learning_rate": 5e-05,
      "loss": 0.9305,
      "step": 2
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.75,
      "eval_auc": 1.0,
      "eval_f1": 0.8,
      "eval_loss": 0.6509360074996948,
      "eval_precision": 0.6666666666666666,
      "eval_recall": 1.0,
      "eval_runtime": 0.2708,
      "eval_samples_per_second": 14.768,
      "eval_steps_per_second": 3.692,
      "step": 2
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.060440063476562,
      "learning_rate": 4.722222222222222e-05,
      "loss": 0.5818,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 19.44124412536621,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.8322,
      "step": 4
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.75,
      "eval_auc": 1.0,
      "eval_f1": 0.8,
      "eval_loss": 0.6616239547729492,
      "eval_precision": 0.6666666666666666,
      "eval_recall": 1.0,
      "eval_runtime": 0.2591,
      "eval_samples_per_second": 15.437,
      "eval_steps_per_second": 3.859,
      "step": 4
    },
    {
      "epoch": 2.5,
      "grad_norm": 12.236566543579102,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.6524,
      "step": 5
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.381006240844727,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.7136,
      "step": 6
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.75,
      "eval_auc": 1.0,
      "eval_f1": 0.8,
      "eval_loss": 0.6655420064926147,
      "eval_precision": 0.6666666666666666,
      "eval_recall": 1.0,
      "eval_runtime": 0.2607,
      "eval_samples_per_second": 15.345,
      "eval_steps_per_second": 3.836,
      "step": 6
    },
    {
      "epoch": 3.5,
      "grad_norm": 12.948503494262695,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.6647,
      "step": 7
    },
    {
      "epoch": 4.0,
      "grad_norm": 15.256901741027832,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.575,
      "step": 8
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5,
      "eval_auc": 0.75,
      "eval_f1": 0.6666666666666666,
      "eval_loss": 0.7162492871284485,
      "eval_precision": 0.5,
      "eval_recall": 1.0,
      "eval_runtime": 0.3132,
      "eval_samples_per_second": 12.77,
      "eval_steps_per_second": 3.192,
      "step": 8
    },
    {
      "epoch": 4.5,
      "grad_norm": 9.987767219543457,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.4975,
      "step": 9
    },
    {
      "epoch": 5.0,
      "grad_norm": 15.961681365966797,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6147,
      "step": 10
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.75,
      "eval_auc": 1.0,
      "eval_f1": 0.8,
      "eval_loss": 0.5041142702102661,
      "eval_precision": 0.6666666666666666,
      "eval_recall": 1.0,
      "eval_runtime": 0.2646,
      "eval_samples_per_second": 15.117,
      "eval_steps_per_second": 3.779,
      "step": 10
    },
    {
      "epoch": 5.5,
      "grad_norm": 8.158454895019531,
      "learning_rate": 2.5e-05,
      "loss": 0.3838,
      "step": 11
    },
    {
      "epoch": 6.0,
      "grad_norm": 11.704643249511719,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.4468,
      "step": 12
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.75,
      "eval_auc": 1.0,
      "eval_f1": 0.8,
      "eval_loss": 0.4621618390083313,
      "eval_precision": 0.6666666666666666,
      "eval_recall": 1.0,
      "eval_runtime": 0.256,
      "eval_samples_per_second": 15.624,
      "eval_steps_per_second": 3.906,
      "step": 12
    },
    {
      "epoch": 6.5,
      "grad_norm": 7.906495571136475,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.3469,
      "step": 13
    },
    {
      "epoch": 7.0,
      "grad_norm": 17.341285705566406,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.684,
      "step": 14
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.75,
      "eval_auc": 1.0,
      "eval_f1": 0.8,
      "eval_loss": 0.5466539859771729,
      "eval_precision": 0.6666666666666666,
      "eval_recall": 1.0,
      "eval_runtime": 0.2641,
      "eval_samples_per_second": 15.146,
      "eval_steps_per_second": 3.786,
      "step": 14
    },
    {
      "epoch": 7.5,
      "grad_norm": 8.669549942016602,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.412,
      "step": 15
    },
    {
      "epoch": 8.0,
      "grad_norm": 13.063695907592773,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.4725,
      "step": 16
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 1.0,
      "eval_auc": 1.0,
      "eval_f1": 1.0,
      "eval_loss": 0.39871343970298767,
      "eval_precision": 1.0,
      "eval_recall": 1.0,
      "eval_runtime": 0.2689,
      "eval_samples_per_second": 14.873,
      "eval_steps_per_second": 3.718,
      "step": 16
    }
  ],
  "logging_steps": 1,
  "max_steps": 20,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.01
      },
      "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": 252135207862272.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}