---
dataset_info:
  features:
  - name: questions
    sequence: string
  - name: question_ids
    sequence: int64
  - name: answers
    sequence:
      sequence: string
  - name: image
    dtype: image
  - name: image_id
    dtype: string
  - name: image_cluster
    dtype: int64
  - name: subquestions
    list:
    - name: answer
      dtype: string
    - name: question
      dtype: string
  - name: base_question
    dtype: string
  - name: base_answer
    sequence: string
  splits:
  - name: train
    num_bytes: 226812706.0
    num_examples: 2000
  download_size: 225548108
  dataset_size: 226812706.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---