metadata
dataset_info:
features:
- name: image
dtype: image
- name: ' question'
dtype: string
- name: ' multiple_choice_answer'
dtype: string
splits:
- name: train
num_bytes: 24565916
num_examples: 37
download_size: 21827866
dataset_size: 24565916
configs:
- config_name: default
data_files:
- split: train
path: data/train-*