metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
sequence: string
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 36050209471
num_examples: 89696
- name: valid
num_bytes: 4504052486
num_examples: 11212
- name: test
num_bytes: 4507640187
num_examples: 11212
download_size: 45062600190
dataset_size: 45061902144
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*
- split: test
path: data/test-*