metadata
dataset_info:
features:
- name: image
dtype: image
- name: question
dtype: string
- name: answer
dtype: string
- name: split
dtype: string
splits:
- name: train
num_bytes: 55685724
num_examples: 249
- name: validation
num_bytes: 12679972
num_examples: 55
download_size: 61822173
dataset_size: 68365696
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*