metadata
dataset_info:
features:
- name: image
dtype: image
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 52064577
num_examples: 10
download_size: 52069802
dataset_size: 52064577
configs:
- config_name: default
data_files:
- split: train
path: data/train-*