test_metrics.json

{
  "canonical_answer_nll": 1.252042673200658,
  "canonical_answer_perplexity": 3.4974798741948536,
  "evaluated_prompts": 2477,
  "generation_rouge_l_f1": 0.3024088364317219,
  "generation_token_f1": 0.42547394221365986,
  "path_samples_per_prompt": 4,
  "preparation": {
    "discarded_prompts": 111,
    "discarded_responses": 111,
    "kept_prompts": 2477
  },
  "reading_logprob_std": 1.261352655227725,
  "reading_span_available_prompts": 2477,
  "reading_span_unavailable_prompts": 0,
  "tokenization_path_logprob_std": 114.96431571929517,
  "trajectory_balance_rmse": 3209.4460347700096
}