metadata
dataset_info:
features:
- name: image
dtype: image
- name: ground_truth
dtype: string
splits:
- name: train
num_bytes: 162409940
num_examples: 850
- name: validation
num_bytes: 19385514
num_examples: 100
- name: test
num_bytes: 9632673
num_examples: 50
download_size: 3361136
dataset_size: 191428127
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*