metadata
dataset_info:
features:
- name: question_id
dtype: string
- name: image
dtype: image
- name: question
dtype: string
- name: answers
sequence: string
- name: category
dtype: string
splits:
- name: val
num_bytes: 2097998373
num_examples: 4319
download_size: 2079513742
dataset_size: 2097998373