metadata
dataset_info:
features:
- name: left_image
dtype: image
- name: right_image
dtype: image
splits:
- name: train
num_bytes: 2229796177
num_examples: 800
- name: validation
num_bytes: 306812020
num_examples: 112
- name: test
num_bytes: 305677512
num_examples: 112
download_size: 2842451111
dataset_size: 2842285709
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*