metadata
dataset_info:
features:
- name: image
dtype: image
- name: caption
dtype: string
- name: conditioning_image
dtype: image
splits:
- name: train
num_bytes: 17121206254.856
num_examples: 53072
- name: test
num_bytes: 641710223
num_examples: 2000
download_size: 17749813525
dataset_size: 17762916477.856003
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*