metadata
dataset_info:
features:
- name: image
sequence:
sequence:
sequence: uint8
- name: mask
sequence:
sequence: float32
splits:
- name: train
num_bytes: 3318193600
num_examples: 100
download_size: 219414034
dataset_size: 3318193600
configs:
- config_name: default
data_files:
- split: train
path: data/train-*