run-3/checkpoint-1532/trainer_state.json

{
  "best_metric": 0.7818003913894325,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-3/checkpoint-1532",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1532,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.53190279006958,
      "learning_rate": 0.00016492442050886904,
      "loss": 0.4371,
      "step": 766
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7475538160469667,
      "eval_f1": 0.7465618860510805,
      "eval_loss": 0.4075860381126404,
      "eval_precision": 0.7495069033530573,
      "eval_recall": 0.7436399217221135,
      "eval_runtime": 132.2506,
      "eval_samples_per_second": 7.728,
      "eval_steps_per_second": 1.936,
      "step": 766
    },
    {
      "epoch": 2.0,
      "grad_norm": 15.610742568969727,
      "learning_rate": 0.00014430886794526042,
      "loss": 0.4097,
      "step": 1532
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7818003913894325,
      "eval_f1": 0.81053525913339,
      "eval_loss": 0.3935793936252594,
      "eval_precision": 0.7162162162162162,
      "eval_recall": 0.9334637964774951,
      "eval_runtime": 132.8723,
      "eval_samples_per_second": 7.692,
      "eval_steps_per_second": 1.927,
      "step": 1532
    }
  ],
  "logging_steps": 500,
  "max_steps": 6894,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6411063430664922,
    "learning_rate": 0.00018553997307247768,
    "num_train_epochs": 9,
    "temperature": 23
  }
}