run-4/checkpoint-192/trainer_state.json

{
  "best_metric": 0.5,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-4/checkpoint-96",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 192,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.6504594683647156,
      "learning_rate": 7.787665972957169e-06,
      "loss": 0.59,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.0,
      "eval_loss": 0.5820541977882385,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 28.7531,
      "eval_samples_per_second": 35.544,
      "eval_steps_per_second": 1.113,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.676121711730957,
      "learning_rate": 0.0,
      "loss": 0.5817,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.0,
      "eval_loss": 0.5805695652961731,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 29.5026,
      "eval_samples_per_second": 34.641,
      "eval_steps_per_second": 1.085,
      "step": 192
    }
  ],
  "logging_steps": 500,
  "max_steps": 192,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7979164056999755,
    "learning_rate": 1.5575331945914338e-05,
    "num_train_epochs": 2,
    "temperature": 5
  }
}