metadata
dataset_info:
features:
- name: ground_truth
dtype: string
- name: image
dtype: image
splits:
- name: train
num_bytes: 250779738
num_examples: 591
- name: validation
num_bytes: 35802053
num_examples: 69
- name: test
num_bytes: 16474058
num_examples: 36
download_size: 302849614
dataset_size: 303055849
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*