metadata
dataset_info:
features:
- name: img_id
dtype: string
- name: original_image
dtype: image
- name: target_image
dtype: image
- name: object_image
dtype: image
splits:
- name: train
num_bytes: 5278330771.798
num_examples: 108093
download_size: 5617667931
dataset_size: 5278330771.798
configs:
- config_name: default
data_files:
- split: train
path: data/train-*