metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 11647824202.89
num_examples: 96915
download_size: 11097789980
dataset_size: 11647824202.89