metadata
dataset_info:
features:
- name: image
dtype: image
- name: ground_truth
dtype: string
splits:
- name: train
num_bytes: 209902708
num_examples: 340
- name: val
num_bytes: 52173333
num_examples: 85
download_size: 262000465
dataset_size: 262076041
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*