metadata
dataset_info:
features:
- name: image
dtype: image
- name: qas
list:
- name: question
dtype: string
- name: answers
list: string
splits:
- name: train
num_bytes: 1929603786.439
num_examples: 12623
download_size: 1926202204
dataset_size: 1929603786.439
configs:
- config_name: default
data_files:
- split: train
path: data/train-*