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: 4092990035.329
num_examples: 42437
download_size: 4072045955
dataset_size: 4092990035.329
configs:
- config_name: default
data_files:
- split: train
path: data/train-*