metadata
dataset_info:
features:
- name: original_image
dtype: image
- name: condtioning_image
dtype: image
- name: overlaid
dtype: image
- name: caption
dtype: string
- name: conditioning_image
dtype: image
splits:
- name: train
num_bytes: 6433410561.125
num_examples: 7519
download_size: 6432551362
dataset_size: 6433410561.125
configs:
- config_name: default
data_files:
- split: train
path: data/train-*