metadata
dataset_info:
features:
- name: original_img
dtype: image
- name: object_name
dtype: string
- name: mask_img
dtype: image
splits:
- name: train
num_bytes: 1102052
num_examples: 8
download_size: 915663
dataset_size: 1102052
configs:
- config_name: default
data_files:
- split: train
path: data/train-*