{
  "best_metric": 0.7439953134153485,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-14/checkpoint-1284",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1284,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 4.230797290802002,
      "learning_rate": 0.0004542691429405582,
      "loss": 0.5244,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7065026362038664,
      "eval_f1": 0.3634053367217281,
      "eval_loss": 0.49707961082458496,
      "eval_mcc": 0.26187368637682734,
      "eval_precision": 0.6559633027522935,
      "eval_recall": 0.2513181019332162,
      "eval_runtime": 3.1642,
      "eval_samples_per_second": 539.474,
      "eval_steps_per_second": 17.066,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.7037988901138306,
      "learning_rate": 0.0004303602406805288,
      "loss": 0.4856,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7047451669595782,
      "eval_f1": 0.5019762845849802,
      "eval_loss": 0.4853743612766266,
      "eval_mcc": 0.3014395863411175,
      "eval_precision": 0.5733634311512416,
      "eval_recall": 0.44639718804920914,
      "eval_runtime": 3.2868,
      "eval_samples_per_second": 519.356,
      "eval_steps_per_second": 16.43,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7830451726913452,
      "learning_rate": 0.0004064513384204994,
      "loss": 0.4758,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7018160515524312,
      "eval_f1": 0.25909752547307136,
      "eval_loss": 0.4858837425708771,
      "eval_mcc": 0.24331468344161114,
      "eval_precision": 0.7542372881355932,
      "eval_recall": 0.15641476274165203,
      "eval_runtime": 3.1648,
      "eval_samples_per_second": 539.369,
      "eval_steps_per_second": 17.063,
      "step": 642
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.6830062866210938,
      "learning_rate": 0.0003825424361604701,
      "loss": 0.4724,
      "step": 856
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7275922671353251,
      "eval_f1": 0.4015444015444016,
      "eval_loss": 0.47459593415260315,
      "eval_mcc": 0.32925006262083517,
      "eval_precision": 0.75,
      "eval_recall": 0.2741652021089631,
      "eval_runtime": 3.7803,
      "eval_samples_per_second": 451.553,
      "eval_steps_per_second": 14.285,
      "step": 856
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.760202407836914,
      "learning_rate": 0.0003586335339004407,
      "loss": 0.4614,
      "step": 1070
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7340363210310487,
      "eval_f1": 0.4733178654292344,
      "eval_loss": 0.4721404016017914,
      "eval_mcc": 0.35044282313920905,
      "eval_precision": 0.6962457337883959,
      "eval_recall": 0.3585237258347979,
      "eval_runtime": 3.1258,
      "eval_samples_per_second": 546.093,
      "eval_steps_per_second": 17.275,
      "step": 1070
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.5421228408813477,
      "learning_rate": 0.0003347246316404113,
      "loss": 0.4617,
      "step": 1284
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7439953134153485,
      "eval_f1": 0.5346112886048987,
      "eval_loss": 0.4643152058124542,
      "eval_mcc": 0.38505007336259955,
      "eval_precision": 0.6783783783783783,
      "eval_recall": 0.44112478031634444,
      "eval_runtime": 3.278,
      "eval_samples_per_second": 520.745,
      "eval_steps_per_second": 16.473,
      "step": 1284
    }
  ],
  "logging_steps": 500,
  "max_steps": 4280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 3148653986640.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8544594605661776,
    "learning_rate": 0.0004781780452005876,
    "num_train_epochs": 20,
    "temperature": 15
  }
}