metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: val
path: data/val-*
dataset_info:
features:
- name: imageA
dtype: image
- name: imageB
dtype: image
- name: label
dtype: image
splits:
- name: train
num_bytes: 3393267984
num_examples: 12000
- name: test
num_bytes: 1196988392
num_examples: 4000
- name: val
num_bytes: 1164865940
num_examples: 4000
download_size: 5814133284
dataset_size: 5755122316