metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype:
class_label:
names:
'0': object_image
'1': original_image
'2': target_image
splits:
- name: train
num_bytes: 795680313.664
num_examples: 127312
download_size: 4071257308
dataset_size: 795680313.664
configs:
- config_name: default
data_files:
- split: train
path: data/train-*