run-4/checkpoint-3176/trainer_state.json

{
  "best_metric": 0.6521328506217535,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-4/checkpoint-1588",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 3176,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.15466690063476562,
      "learning_rate": 0.00028711293555461854,
      "loss": 0.047,
      "step": 1588
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6521328506217535,
      "eval_f1": 0.006741573033707866,
      "eval_loss": 0.043082475662231445,
      "eval_mcc": -0.07334683954324206,
      "eval_precision": 0.06976744186046512,
      "eval_recall": 0.0035419126328217238,
      "eval_runtime": 23.3677,
      "eval_samples_per_second": 543.742,
      "eval_steps_per_second": 17.032,
      "step": 1588
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2909219563007355,
      "learning_rate": 0.0002153347016659639,
      "loss": 0.0437,
      "step": 3176
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6499291673225248,
      "eval_f1": 0.010676156583629894,
      "eval_loss": 0.04202444478869438,
      "eval_mcc": -0.07414646393220078,
      "eval_precision": 0.09195402298850575,
      "eval_recall": 0.005667060212514758,
      "eval_runtime": 23.3008,
      "eval_samples_per_second": 545.304,
      "eval_steps_per_second": 17.081,
      "step": 3176
    }
  ],
  "logging_steps": 500,
  "max_steps": 7940,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 7817788152600.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.05479484259344969,
    "learning_rate": 0.0003588911694432732,
    "num_train_epochs": 5,
    "temperature": 13
  }
}