metadata
dataset_info:
features:
- name: ground_truth
dtype: string
- name: image
dtype: image
splits:
- name: train
num_bytes: 608236194
num_examples: 1011
- name: validation
num_bytes: 76008447
num_examples: 113
download_size: 684027014
dataset_size: 684244641
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*