metadata
dataset_info:
features:
- name: images
dtype: image
- name: label
sequence: int32
splits:
- name: train
num_bytes: 75545704
num_examples: 89400
download_size: 42418183
dataset_size: 75545704
configs:
- config_name: default
data_files:
- split: train
path: data/train-*