metadata
dataset_info:
features:
- name: image
dtype: image
- name: masked_image
dtype: image
- name: mask
sequence:
sequence: int64
splits:
- name: train
num_bytes: 334098735
num_examples: 54
download_size: 45886126
dataset_size: 334098735
configs:
- config_name: default
data_files:
- split: train
path: data/train-*