metadata
dataset_info:
features:
- name: image
dtype: image
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 1074552713.904604
num_examples: 1928
- name: test
num_bytes: 269195519.0953961
num_examples: 483
download_size: 1343698644
dataset_size: 1343748233
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*