{
  "best_metric": 0.5762376237623762,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-3/checkpoint-1782",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2079,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.8068264722824097,
      "learning_rate": 1.8960352341732743e-05,
      "loss": 0.5426,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.500990099009901,
      "eval_loss": 0.5363946557044983,
      "eval_runtime": 55.6857,
      "eval_samples_per_second": 9.069,
      "eval_steps_per_second": 0.287,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7480373978614807,
      "learning_rate": 1.580029361811062e-05,
      "loss": 0.5354,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5564356435643565,
      "eval_loss": 0.5311887264251709,
      "eval_runtime": 55.3283,
      "eval_samples_per_second": 9.127,
      "eval_steps_per_second": 0.289,
      "step": 594
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.3350061178207397,
      "learning_rate": 1.2640234894488496e-05,
      "loss": 0.5312,
      "step": 891
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5603960396039604,
      "eval_loss": 0.5280594229698181,
      "eval_runtime": 55.2512,
      "eval_samples_per_second": 9.14,
      "eval_steps_per_second": 0.29,
      "step": 891
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1148613691329956,
      "learning_rate": 9.480176170866372e-06,
      "loss": 0.5287,
      "step": 1188
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5643564356435643,
      "eval_loss": 0.5284162759780884,
      "eval_runtime": 55.166,
      "eval_samples_per_second": 9.154,
      "eval_steps_per_second": 0.29,
      "step": 1188
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.2209173440933228,
      "learning_rate": 6.320117447244248e-06,
      "loss": 0.5278,
      "step": 1485
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5683168316831683,
      "eval_loss": 0.5252435803413391,
      "eval_runtime": 55.1513,
      "eval_samples_per_second": 9.157,
      "eval_steps_per_second": 0.29,
      "step": 1485
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.3163679838180542,
      "learning_rate": 3.160058723622124e-06,
      "loss": 0.5264,
      "step": 1782
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5762376237623762,
      "eval_loss": 0.525602400302887,
      "eval_runtime": 57.497,
      "eval_samples_per_second": 8.783,
      "eval_steps_per_second": 0.278,
      "step": 1782
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.077811360359192,
      "learning_rate": 0.0,
      "loss": 0.5257,
      "step": 2079
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5702970297029702,
      "eval_loss": 0.524278998374939,
      "eval_runtime": 55.3503,
      "eval_samples_per_second": 9.124,
      "eval_steps_per_second": 0.289,
      "step": 2079
    }
  ],
  "logging_steps": 500,
  "max_steps": 2079,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 5444902981980.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7228353037734928,
    "learning_rate": 2.212041106535487e-05,
    "num_train_epochs": 7,
    "temperature": 7
  }
}