metadata
dataset_info:
features:
- name: img1
dtype: string
- name: img2
dtype: string
- name: label
dtype: int64
splits:
- name: train
num_bytes: 3559603
num_examples: 60024
- name: test
num_bytes: 249637
num_examples: 4270
download_size: 1956946
dataset_size: 3809240
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*