metadata
dataset_info:
features:
- name: path
dtype: string
- name: label
sequence: uint8
- name: image
dtype: image
splits:
- name: train
num_bytes: 535025564.875
num_examples: 34561
- name: validation
num_bytes: 104554021.625
num_examples: 6491
download_size: 638557710
dataset_size: 639579586.5
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*