README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: string
    - name: title
      dtype: string
    - name: context
      dtype: string
    - name: question
      dtype: string
    - name: answers
      sequence:
        - name: text
          dtype: string
        - name: answer_start
          dtype: int32
  splits:
    - name: train
      num_bytes: 79075277.43924017
      num_examples: 87300
    - name: validation
      num_bytes: 10390745.809649954
      num_examples: 10487
  download_size: 16144193
  dataset_size: 89466023.24889013
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*