---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: question
    dtype: string
  - name: paragraphs
    sequence: int64
  - name: relevant
    dtype: int64
  - name: answer
    struct:
    - name: start
      dtype: int64
    - name: text
      dtype: string
  splits:
  - name: train
    num_bytes: 3666495
    num_examples: 21714
  - name: validation
    num_bytes: 521099
    num_examples: 3009
  download_size: 3016668
  dataset_size: 4187594
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---