metadata
dataset_info:
features:
- name: image
dtype: binary
splits:
- name: train
num_bytes: 71624856
num_examples: 3670
download_size: 71655918
dataset_size: 71624856
configs:
- config_name: default
data_files:
- split: train
path: data/train-*