---
dataset_info:
  features:
  - name: question
    sequence: string
  - name: label
    sequence: int64
  - name: image
    dtype: image
  - name: sub_questions
    list:
    - name: answer
      dtype: int64
    - name: question
      dtype: string
  - name: base_question
    dtype: string
  - name: base_answer
    dtype: int64
  splits:
  - name: train
    num_bytes: 33411773.0
    num_examples: 250
  download_size: 33258535
  dataset_size: 33411773.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---