metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 2473559738
num_examples: 785
download_size: 2473661020
dataset_size: 2473559738
configs:
- config_name: default
data_files:
- split: train
path: data/train-*