---
dataset_info:
  features:
  - name: document_id
    dtype: int32
  - name: question
    dtype: string
  - name: id
    dtype: int32
  - name: answers
    sequence: string
  - name: chunks
    sequence: string
  - name: correct_chunks
    sequence: int64
  - name: num_words
    dtype: int64
  - name: num_tokens
    dtype: int64
  splits:
  - name: train
    num_bytes: 43249481.2332838
    num_examples: 1335
  download_size: 5047983
  dataset_size: 43249481.2332838
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---