---
dataset_info:
  features:
  - name: question
    dtype: string
  - name: ground_truth
    dtype: float64
  - name: generated_answers
    list: string
  - name: num_correct_answers
    dtype: int64
  - name: num_generations
    dtype: int64
  - name: centroid_diversity_scores
    list: float64
  - name: avg_centroid_diversity_score
    dtype: float64
  - name: avg_pairwise_diversity
    dtype: float64
  - name: max_pairwise_diversity
    dtype: float64
  - name: var_pairwise_diversity
    dtype: float64
  splits:
  - name: train
    num_bytes: 5546725
    num_examples: 1000
  download_size: 5410896
  dataset_size: 5546725
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---