metadata
dataset_info:
features:
- name: image
dtype: image
- name: ground_truth
dtype: string
splits:
- name: train
num_bytes: 4340493430.975
num_examples: 9995
- name: validation
num_bytes: 1808061
num_examples: 5
download_size: 3469197006
dataset_size: 4342301491.975
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*