run-11/checkpoint-428/trainer_state.json

{
  "best_metric": 0.7311072056239016,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-11/checkpoint-428",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 428,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 7.572170734405518,
      "learning_rate": 0.00012123385931636098,
      "loss": 0.5934,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7082601054481547,
      "eval_f1": 0.435374149659864,
      "eval_loss": 0.5416340827941895,
      "eval_mcc": 0.28153811271695595,
      "eval_precision": 0.6134185303514377,
      "eval_recall": 0.3374340949033392,
      "eval_runtime": 3.1348,
      "eval_samples_per_second": 544.534,
      "eval_steps_per_second": 17.226,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.894651412963867,
      "learning_rate": 0.00010607962690181587,
      "loss": 0.5286,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7311072056239016,
      "eval_f1": 0.5069817400644469,
      "eval_loss": 0.5141445398330688,
      "eval_mcc": 0.35062725647270626,
      "eval_precision": 0.6519337016574586,
      "eval_recall": 0.4147627416520211,
      "eval_runtime": 3.1601,
      "eval_samples_per_second": 540.173,
      "eval_steps_per_second": 17.088,
      "step": 428
    }
  ],
  "logging_steps": 500,
  "max_steps": 1926,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 1049551328880.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9696180923551023,
    "learning_rate": 0.0001363880917309061,
    "num_train_epochs": 9,
    "temperature": 20
  }
}