run-30/checkpoint-112/trainer_state.json

{
  "best_metric": 0.8062622309197651,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-30/checkpoint-112",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 112,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.488581895828247,
      "learning_rate": 0.00035374495201026426,
      "loss": 0.5645,
      "step": 56
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7847358121330724,
      "eval_f1": 0.8049645390070922,
      "eval_loss": 0.4679928421974182,
      "eval_precision": 0.7358184764991896,
      "eval_recall": 0.8884540117416829,
      "eval_runtime": 28.8427,
      "eval_samples_per_second": 35.434,
      "eval_steps_per_second": 1.109,
      "step": 56
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.869302749633789,
      "learning_rate": 0.00029478746000855357,
      "loss": 0.4598,
      "step": 112
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8062622309197651,
      "eval_f1": 0.8047337278106508,
      "eval_loss": 0.4556478261947632,
      "eval_precision": 0.8111332007952287,
      "eval_recall": 0.7984344422700587,
      "eval_runtime": 28.6037,
      "eval_samples_per_second": 35.73,
      "eval_steps_per_second": 1.119,
      "step": 112
    }
  ],
  "logging_steps": 500,
  "max_steps": 392,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 55,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9318712276013821,
    "learning_rate": 0.000412702444011975,
    "num_train_epochs": 7,
    "per_device_train_batch_size": 55,
    "temperature": 27
  }
}