metadata
dataset_info:
features:
- name: image
dtype: image
- name: mask
dtype: image
- name: split
dtype: string
splits:
- name: train
num_bytes: 34450965
num_examples: 3345
download_size: 249337909
dataset_size: 34450965
configs:
- config_name: default
data_files:
- split: train
path: data/train-*