metadata
dataset_info:
features:
- name: image
dtype: image
- name: conditioning
dtype: image
- name: prompt
sequence: string
splits:
- name: train
num_bytes: 3854203
num_examples: 4
download_size: 3857683
dataset_size: 3854203
configs:
- config_name: default
data_files:
- split: train
path: data/train-*