{
  "best_metric": 0.8111545988258317,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-12/checkpoint-288",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 288,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.4721163511276245,
      "learning_rate": 0.0005951476674473283,
      "loss": 0.4971,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7671232876712328,
      "eval_f1": 0.7989864864864864,
      "eval_loss": 0.4431186616420746,
      "eval_precision": 0.7028231797919762,
      "eval_recall": 0.9256360078277887,
      "eval_runtime": 28.4119,
      "eval_samples_per_second": 35.971,
      "eval_steps_per_second": 1.126,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.945635199546814,
      "learning_rate": 0.00029757383372366414,
      "loss": 0.4421,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7915851272015656,
      "eval_f1": 0.8149435273675064,
      "eval_loss": 0.4262682795524597,
      "eval_precision": 0.7328125,
      "eval_recall": 0.9178082191780822,
      "eval_runtime": 28.1497,
      "eval_samples_per_second": 36.306,
      "eval_steps_per_second": 1.137,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.666189193725586,
      "learning_rate": 0.0,
      "loss": 0.4148,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8111545988258317,
      "eval_f1": 0.8146013448607108,
      "eval_loss": 0.40171727538108826,
      "eval_precision": 0.8,
      "eval_recall": 0.8297455968688845,
      "eval_runtime": 28.233,
      "eval_samples_per_second": 36.199,
      "eval_steps_per_second": 1.133,
      "step": 288
    }
  ],
  "logging_steps": 500,
  "max_steps": 288,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 707085591840.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7419170223133877,
    "learning_rate": 0.0008927215011709925,
    "num_train_epochs": 3,
    "temperature": 30
  }
}