{
  "best_metric": 0.8291284403669725,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-16/checkpoint-1054",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1581,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 15.605107307434082,
      "learning_rate": 0.00011351736809377586,
      "loss": 1.2624,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8256880733944955,
      "eval_loss": 1.1314730644226074,
      "eval_runtime": 2.8139,
      "eval_samples_per_second": 309.89,
      "eval_steps_per_second": 2.488,
      "step": 527
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.443777084350586,
      "learning_rate": 7.567824539585057e-05,
      "loss": 0.6363,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8291284403669725,
      "eval_loss": 1.1333012580871582,
      "eval_runtime": 2.8147,
      "eval_samples_per_second": 309.801,
      "eval_steps_per_second": 2.487,
      "step": 1054
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 3.791092368976196e-05,
      "loss": 0.4911,
      "step": 1581
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8291284403669725,
      "eval_loss": 1.1619139909744263,
      "eval_runtime": 2.8192,
      "eval_samples_per_second": 309.306,
      "eval_steps_per_second": 2.483,
      "step": 1581
    }
  ],
  "logging_steps": 500,
  "max_steps": 2108,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 24228458327580.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.5930583577377427,
    "learning_rate": 0.00015135649079170115,
    "num_train_epochs": 4,
    "temperature": 20
  }
}