metadata
dataset_info:
features:
- name: left
dtype: image
- name: right
dtype: image
splits:
- name: train
num_bytes: 274014841
num_examples: 399
download_size: 274036347
dataset_size: 274014841
configs:
- config_name: default
data_files:
- split: train
path: data/train-*