README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: string
    - name: title
      dtype: string
    - name: context
      dtype: string
    - name: question
      dtype: string
    - name: answers
      struct:
        - name: answer_start
          sequence: int32
        - name: text
          sequence: string
  splits:
    - name: train
      num_bytes: 79301631
      num_examples: 87588
    - name: validation
      num_bytes: 5239631
      num_examples: 5285
    - name: test
      num_bytes: 5233006
      num_examples: 5285
  download_size: 19809326
  dataset_size: 89774268
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*