{
  "best_metric": 0.2855555533152525,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-0/checkpoint-1605",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1605,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 2.1133761405944824,
      "learning_rate": 1.2463453586617104e-06,
      "loss": 0.6018,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.575204074382782,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.971,
      "eval_samples_per_second": 1074.116,
      "eval_steps_per_second": 134.908,
      "step": 535
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 4.266573905944824,
      "learning_rate": 6.82388182796683e-07,
      "loss": 0.5438,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.54793781042099,
      "eval_matthews_correlation": 0.19076704585900842,
      "eval_runtime": 1.3928,
      "eval_samples_per_second": 748.864,
      "eval_steps_per_second": 94.057,
      "step": 1070
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 6.442965507507324,
      "learning_rate": 1.1843100693165573e-07,
      "loss": 0.5106,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.5384307503700256,
      "eval_matthews_correlation": 0.2855555533152525,
      "eval_runtime": 0.9899,
      "eval_samples_per_second": 1053.649,
      "eval_steps_per_second": 132.337,
      "step": 1605
    }
  ],
  "logging_steps": 500,
  "max_steps": 1605,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 128036208935724.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.8103025345267377e-06,
    "num_train_epochs": 3,
    "per_device_train_batch_size": 16,
    "seed": 40
  }
}