{
  "best_metric": 0.7006444053895724,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-7/checkpoint-642",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 642,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 5.4419636726379395,
      "learning_rate": 4.778212463646413e-05,
      "loss": 0.4867,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6649091974223784,
      "eval_f1": 0.01718213058419244,
      "eval_loss": 0.4545574486255646,
      "eval_mcc": 0.009529862152017439,
      "eval_precision": 0.38461538461538464,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.147,
      "eval_samples_per_second": 542.424,
      "eval_steps_per_second": 17.159,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.8326942920684814,
      "learning_rate": 3.9818437197053446e-05,
      "loss": 0.4528,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6959578207381371,
      "eval_f1": 0.40955631399317405,
      "eval_loss": 0.44587329030036926,
      "eval_mcc": 0.24713473389349372,
      "eval_precision": 0.5806451612903226,
      "eval_recall": 0.3163444639718805,
      "eval_runtime": 3.2126,
      "eval_samples_per_second": 531.351,
      "eval_steps_per_second": 16.809,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.2975902557373047,
      "learning_rate": 3.1854749757642754e-05,
      "loss": 0.4424,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7006444053895724,
      "eval_f1": 0.396694214876033,
      "eval_loss": 0.4441916346549988,
      "eval_mcc": 0.25354492903122977,
      "eval_precision": 0.60431654676259,
      "eval_recall": 0.29525483304042177,
      "eval_runtime": 3.1517,
      "eval_samples_per_second": 541.618,
      "eval_steps_per_second": 17.134,
      "step": 642
    }
  ],
  "logging_steps": 500,
  "max_steps": 1498,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 1574326993320.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7570540466770024,
    "learning_rate": 5.574581207587482e-05,
    "num_train_epochs": 7,
    "temperature": 25
  }
}