metadata
dataset_info:
features:
- name: source
dtype: image
- name: prompt
dtype: string
- name: target
dtype: image
splits:
- name: train
num_bytes: 1907395
num_examples: 9
download_size: 639510
dataset_size: 1907395
configs:
- config_name: default
data_files:
- split: train
path: data/train-*