metadata
dataset_info:
features:
- name: source_image
dtype: image
- name: condition_img
dtype: image
- name: caption
dtype: string
splits:
- name: train
num_bytes: 8268336911.5
num_examples: 19842
download_size: 8266042780
dataset_size: 8268336911.5
configs:
- config_name: default
data_files:
- split: train
path: data/train-*