run-6/checkpoint-100/trainer_state.json

{
  "best_metric": 0.6154598825831703,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-6/checkpoint-100",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.0887755155563354,
      "learning_rate": 0.0002701514784814016,
      "loss": 0.3405,
      "step": 50
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5440313111545988,
      "eval_f1": 0.21548821548821548,
      "eval_loss": 0.3118150532245636,
      "eval_precision": 0.7710843373493976,
      "eval_recall": 0.12524461839530332,
      "eval_runtime": 28.2208,
      "eval_samples_per_second": 36.214,
      "eval_steps_per_second": 1.134,
      "step": 50
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1933519840240479,
      "learning_rate": 0.00023155841012691565,
      "loss": 0.3081,
      "step": 100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6154598825831703,
      "eval_f1": 0.43776824034334766,
      "eval_loss": 0.2966625690460205,
      "eval_precision": 0.8138297872340425,
      "eval_recall": 0.299412915851272,
      "eval_runtime": 28.3705,
      "eval_samples_per_second": 36.023,
      "eval_steps_per_second": 1.128,
      "step": 100
    }
  ],
  "logging_steps": 500,
  "max_steps": 400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 62,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.43186153278710804,
    "learning_rate": 0.0003087445468358875,
    "num_train_epochs": 8,
    "per_device_train_batch_size": 62,
    "temperature": 10
  }
}