README.md

metadata

dataset_info:
  features:
    - name: question
      dtype: string
    - name: answer
      dtype: string
    - name: reasoning_type
      dtype: string
    - name: source
      dtype: string
    - name: section_id
      dtype: string
    - name: query_id
      dtype: string
    - name: passage
      dtype: string
    - name: answers_spans
      struct:
        - name: spans
          list: string
        - name: types
          list: string
    - name: id
      dtype: string
    - name: question_concept
      dtype: string
    - name: choices
      struct:
        - name: label
          list: string
        - name: text
          list: string
    - name: answerKey
      dtype: string
  splits:
    - name: train
      num_bytes: 18305243
      num_examples: 8000
    - name: validation
      num_bytes: 1193287
      num_examples: 1000
    - name: test
      num_bytes: 1209066
      num_examples: 1000
  download_size: 11305044
  dataset_size: 20707596
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
task_categories:
  - question-answering
language:
  - en
size_categories:
  - 1K<n<10K