---
dataset_info:
  features:
  - name: question
    dtype: string
  - name: answers
    list: string
  - name: relation
    dtype: string
  - name: subject
    dtype: string
  - name: entity_frequency
    dtype: int64
  - name: retrieved_docs
    list:
    - name: doc_id
      dtype: string
    - name: rank
      dtype: int64
    - name: rerank_score
      dtype: float64
    - name: rrf
      dtype: float64
    - name: score
      dtype: float64
    - name: source
      dtype: string
    - name: text
      dtype: string
    - name: title
      dtype: string
  - name: hit_rank
    dtype: int64
  - name: bucket
    dtype: string
  splits:
  - name: train
    num_bytes: 345190054
    num_examples: 15250
  download_size: 195450365
  dataset_size: 345190054
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---