metadata
dataset_info:
features:
- name: __key__
dtype: string
- name: __url__
dtype: string
- name: image
dtype: image
- name: labels
dtype: int64
splits:
- name: train
num_bytes: 2911189505.25
num_examples: 35126
- name: test
num_bytes: 3537477981.25
num_examples: 42670
download_size: 6440769124
dataset_size: 6448667486.5
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*