metadata
dataset_info:
features:
- name: image1
dtype: string
- name: image2
dtype: string
- name: label
dtype: int64
splits:
- name: train
num_bytes: 920973
num_examples: 5166
download_size: 80188
dataset_size: 920973
configs:
- config_name: default
data_files:
- split: train
path: data/train-*