run-6/checkpoint-428/trainer_state.json

{
  "best_metric": 0.7135325131810193,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-6/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": 4.246974468231201,
      "learning_rate": 0.00031328826443341315,
      "loss": 0.5126,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7006444053895724,
      "eval_f1": 0.31039136302294196,
      "eval_loss": 0.494042307138443,
      "eval_mcc": 0.23807439640980294,
      "eval_precision": 0.6686046511627907,
      "eval_recall": 0.20210896309314588,
      "eval_runtime": 3.1371,
      "eval_samples_per_second": 544.127,
      "eval_steps_per_second": 17.213,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.0712907314300537,
      "learning_rate": 0.0,
      "loss": 0.4771,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7135325131810193,
      "eval_f1": 0.43205574912891986,
      "eval_loss": 0.47912341356277466,
      "eval_mcc": 0.29261527491603634,
      "eval_precision": 0.636986301369863,
      "eval_recall": 0.3268892794376098,
      "eval_runtime": 3.2343,
      "eval_samples_per_second": 527.774,
      "eval_steps_per_second": 16.696,
      "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": 0.8460079304212,
    "learning_rate": 0.0006265765288668263,
    "num_train_epochs": 2,
    "temperature": 22
  }
}