{
  "best_metric": 0.47624343628277643,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-3/checkpoint-1072",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1340,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 0.5160695314407349,
      "eval_matthews_correlation": 0.3884569188624487,
      "eval_runtime": 0.688,
      "eval_samples_per_second": 1516.05,
      "eval_steps_per_second": 95.934,
      "step": 268
    },
    {
      "epoch": 1.87,
      "learning_rate": 4.7623518427226e-06,
      "loss": 0.5077,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.4879527986049652,
      "eval_matthews_correlation": 0.4537159672261629,
      "eval_runtime": 1.8327,
      "eval_samples_per_second": 569.098,
      "eval_steps_per_second": 36.012,
      "step": 536
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.49178388714790344,
      "eval_matthews_correlation": 0.4752024043308134,
      "eval_runtime": 0.7459,
      "eval_samples_per_second": 1398.283,
      "eval_steps_per_second": 88.482,
      "step": 804
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.9276186030067667e-06,
      "loss": 0.3683,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.5047063231468201,
      "eval_matthews_correlation": 0.47624343628277643,
      "eval_runtime": 1.64,
      "eval_samples_per_second": 635.968,
      "eval_steps_per_second": 40.243,
      "step": 1072
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.5170151591300964,
      "eval_matthews_correlation": 0.4618671668713098,
      "eval_runtime": 1.1213,
      "eval_samples_per_second": 930.176,
      "eval_steps_per_second": 58.861,
      "step": 1340
    }
  ],
  "logging_steps": 500,
  "max_steps": 1340,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 190971935742888.0,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 7.597085082438434e-06,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 32,
    "seed": 25
  }
}