README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: string
    - name: title
      dtype: string
    - name: context
      dtype: string
    - name: questions
      sequence: string
    - name: type
      dtype: string
  splits:
    - name: train
      num_bytes: 367983987
      num_examples: 197698
    - name: validation
      num_bytes: 66781868
      num_examples: 37042
  download_size: 229759911
  dataset_size: 434765855
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*

행정 문서 대상 기계독해 데이터