metadata
dataset_info:
features:
- name: image
dtype: image
- name: conditioning_image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 3802768696
num_examples: 45000
- name: validation
num_bytes: 170783650
num_examples: 2000
download_size: 3929429292
dataset_size: 3973552346
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*