---
dataset_info:
  features:
  - name: custom_id
    dtype: string
  - name: author_info
    struct:
    - name: email_address
      dtype: string
  - name: question_categories
    sequence: string
  - name: image_attribution
    dtype: string
  - name: question
    dtype: string
  - name: question_images
    sequence: image
  - name: final_answer
    dtype: string
  - name: rationale_text
    dtype: string
  - name: rationale_images
    sequence: image
  splits:
  - name: train
    num_bytes: 33467517.0
    num_examples: 87
  - name: test
    num_bytes: 78533251.0
    num_examples: 203
  download_size: 38575645117
  dataset_size: 112000768.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---