run-21/checkpoint-428/trainer_state.json

{
  "best_metric": 0.7234915055653193,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-21/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": 7.830043792724609,
      "learning_rate": 9.118416079790422e-05,
      "loss": 0.5332,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.705916813122437,
      "eval_f1": 0.49087221095334677,
      "eval_loss": 0.4951828718185425,
      "eval_mcc": 0.2979069251542787,
      "eval_precision": 0.580335731414868,
      "eval_recall": 0.4253075571177504,
      "eval_runtime": 3.1297,
      "eval_samples_per_second": 545.419,
      "eval_steps_per_second": 17.254,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.675640106201172,
      "learning_rate": 0.0,
      "loss": 0.4845,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7234915055653193,
      "eval_f1": 0.51440329218107,
      "eval_loss": 0.4795408844947815,
      "eval_mcc": 0.3384723199828929,
      "eval_precision": 0.6203473945409429,
      "eval_recall": 0.43936731107205623,
      "eval_runtime": 3.1418,
      "eval_samples_per_second": 543.315,
      "eval_steps_per_second": 17.187,
      "step": 428
    }
  ],
  "logging_steps": 500,
  "max_steps": 428,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 1049551328880.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.864112168223305,
    "learning_rate": 0.00018236832159580844,
    "num_train_epochs": 2,
    "temperature": 19
  }
}