{
  "best_metric": 0.8268101761252447,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-13/checkpoint-216",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 270,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.1670913696289062,
      "learning_rate": 0.0002855230244841258,
      "loss": 0.5191,
      "step": 54
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7407045009784736,
      "eval_f1": 0.7881694644284573,
      "eval_loss": 0.47280439734458923,
      "eval_precision": 0.6662162162162162,
      "eval_recall": 0.9647749510763209,
      "eval_runtime": 29.2616,
      "eval_samples_per_second": 34.926,
      "eval_steps_per_second": 1.094,
      "step": 54
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.121979713439941,
      "learning_rate": 0.00021414226836309432,
      "loss": 0.4574,
      "step": 108
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7798434442270059,
      "eval_f1": 0.7770069375619425,
      "eval_loss": 0.42865216732025146,
      "eval_precision": 0.7871485943775101,
      "eval_recall": 0.7671232876712328,
      "eval_runtime": 28.412,
      "eval_samples_per_second": 35.971,
      "eval_steps_per_second": 1.126,
      "step": 108
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5867382287979126,
      "learning_rate": 0.0001427615122420629,
      "loss": 0.4287,
      "step": 162
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.815068493150685,
      "eval_f1": 0.8205128205128206,
      "eval_loss": 0.40865078568458557,
      "eval_precision": 0.7970479704797048,
      "eval_recall": 0.8454011741682974,
      "eval_runtime": 28.2445,
      "eval_samples_per_second": 36.184,
      "eval_steps_per_second": 1.133,
      "step": 162
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.5582785606384277,
      "learning_rate": 7.138075612103145e-05,
      "loss": 0.4101,
      "step": 216
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8268101761252447,
      "eval_f1": 0.8392370572207084,
      "eval_loss": 0.3997044265270233,
      "eval_precision": 0.7830508474576271,
      "eval_recall": 0.9041095890410958,
      "eval_runtime": 28.2876,
      "eval_samples_per_second": 36.129,
      "eval_steps_per_second": 1.131,
      "step": 216
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.5253950357437134,
      "learning_rate": 0.0,
      "loss": 0.3979,
      "step": 270
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8268101761252447,
      "eval_f1": 0.837763519706691,
      "eval_loss": 0.3981335759162903,
      "eval_precision": 0.7879310344827586,
      "eval_recall": 0.8943248532289628,
      "eval_runtime": 28.8031,
      "eval_samples_per_second": 35.482,
      "eval_steps_per_second": 1.111,
      "step": 270
    }
  ],
  "logging_steps": 500,
  "max_steps": 270,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1178475986400.0,
  "train_batch_size": 57,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7778146133600564,
    "learning_rate": 0.0003569037806051572,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 57,
    "temperature": 18
  }
}