{
  "best_global_step": 3988,
  "best_metric": 0.7874481941977501,
  "best_model_checkpoint": "./deberta_bio_result/checkpoint-3988",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 3988,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25075225677031093,
      "grad_norm": 325735.625,
      "learning_rate": 1.7497492477432297e-05,
      "loss": 0.412,
      "step": 500
    },
    {
      "epoch": 0.5015045135406219,
      "grad_norm": 167717.0,
      "learning_rate": 1.498996990972919e-05,
      "loss": 0.3617,
      "step": 1000
    },
    {
      "epoch": 0.7522567703109327,
      "grad_norm": 278521.625,
      "learning_rate": 1.2482447342026078e-05,
      "loss": 0.3601,
      "step": 1500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.843530591775326,
      "eval_f1": 0.770546056260342,
      "eval_loss": 0.3484754264354706,
      "eval_runtime": 122.603,
      "eval_samples_per_second": 130.111,
      "eval_steps_per_second": 4.07,
      "step": 1994
    },
    {
      "epoch": 1.0030090270812437,
      "grad_norm": 120681.1328125,
      "learning_rate": 9.974924774322969e-06,
      "loss": 0.3392,
      "step": 2000
    },
    {
      "epoch": 1.2537612838515546,
      "grad_norm": 123770.0234375,
      "learning_rate": 7.46740220661986e-06,
      "loss": 0.3112,
      "step": 2500
    },
    {
      "epoch": 1.5045135406218657,
      "grad_norm": 149896.875,
      "learning_rate": 4.959879638916751e-06,
      "loss": 0.3082,
      "step": 3000
    },
    {
      "epoch": 1.7552657973921764,
      "grad_norm": 137855.703125,
      "learning_rate": 2.452357071213641e-06,
      "loss": 0.305,
      "step": 3500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8649699097291875,
      "eval_f1": 0.7874481941977501,
      "eval_loss": 0.33768922090530396,
      "eval_runtime": 122.6162,
      "eval_samples_per_second": 130.097,
      "eval_steps_per_second": 4.07,
      "step": 3988
    }
  ],
  "logging_steps": 500,
  "max_steps": 3988,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8394050978503680.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}