metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 5143949710.7382965
num_examples: 31500
download_size: 5444917886
dataset_size: 5143949710.7382965
configs:
- config_name: default
data_files:
- split: train
path: data/train-*