metadata
dataset_info:
features:
- name: text
dtype: string
- name: image
dtype: image
- name: conditioning_image
dtype: image
splits:
- name: train
num_bytes: 4630391106
num_examples: 28000
download_size: 4494185653
dataset_size: 4630391106
configs:
- config_name: default
data_files:
- split: train
path: data/train-*