metadata
dataset_info:
features:
- name: images
dtype: image
- name: questions
dtype: string
- name: answers
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 79685301.5
num_examples: 1500
download_size: 79102582
dataset_size: 79685301.5
configs:
- config_name: default
data_files:
- split: train
path: data/train-*