metadata
dataset_info:
features:
- name: image
dtype: image
- name: question
dtype: string
- name: choices
sequence: string
- name: answer
dtype: int8
- name: correct_answer
dtype: string
splits:
- name: train
num_bytes: 246244065
num_examples: 3914
- name: validation
num_bytes: 82542203
num_examples: 1328
- name: test
num_bytes: 74700536
num_examples: 1208
download_size: 389785235
dataset_size: 403486804
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*