metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
sequence: float32
splits:
- name: train
num_bytes: 68567616.60800001
num_examples: 25128
download_size: 52987162
dataset_size: 68567616.60800001
configs:
- config_name: default
data_files:
- split: train
path: data/train-*