{
  "best_metric": 0.821917808219178,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-18/checkpoint-576",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 768,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.9128750562667847,
      "learning_rate": 0.000468107445983249,
      "loss": 0.5305,
      "step": 192
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7847358121330724,
      "eval_f1": 0.7920604914933836,
      "eval_loss": 0.45163166522979736,
      "eval_mcc": 0.5708901120719992,
      "eval_precision": 0.7659963436928702,
      "eval_recall": 0.8199608610567515,
      "eval_runtime": 67.0893,
      "eval_samples_per_second": 15.233,
      "eval_steps_per_second": 0.477,
      "step": 192
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.839444160461426,
      "learning_rate": 0.00035108058448743673,
      "loss": 0.4711,
      "step": 384
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7984344422700587,
      "eval_f1": 0.7944111776447106,
      "eval_loss": 0.43096378445625305,
      "eval_mcc": 0.5973265692911504,
      "eval_precision": 0.8105906313645621,
      "eval_recall": 0.7788649706457925,
      "eval_runtime": 66.944,
      "eval_samples_per_second": 15.266,
      "eval_steps_per_second": 0.478,
      "step": 384
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.6563186645507812,
      "learning_rate": 0.0002340537229916245,
      "loss": 0.4287,
      "step": 576
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.821917808219178,
      "eval_f1": 0.8286252354048964,
      "eval_loss": 0.4311530292034149,
      "eval_mcc": 0.6458172559023482,
      "eval_precision": 0.7985480943738656,
      "eval_recall": 0.8610567514677103,
      "eval_runtime": 67.6871,
      "eval_samples_per_second": 15.099,
      "eval_steps_per_second": 0.473,
      "step": 576
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.9438397884368896,
      "learning_rate": 0.00011702686149581226,
      "loss": 0.4132,
      "step": 768
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8170254403131115,
      "eval_f1": 0.8227488151658768,
      "eval_loss": 0.4179845452308655,
      "eval_mcc": 0.6353771789166361,
      "eval_precision": 0.7977941176470589,
      "eval_recall": 0.8493150684931506,
      "eval_runtime": 66.772,
      "eval_samples_per_second": 15.306,
      "eval_steps_per_second": 0.479,
      "step": 768
    }
  ],
  "logging_steps": 500,
  "max_steps": 960,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 942780789120.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8570009057234442,
    "learning_rate": 0.0005851343074790613,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 16,
    "temperature": 44
  }
}