metadata
dataset_info:
features:
- name: image_filename
dtype: string
- name: query
dtype: string
- name: image
dtype: image
splits:
- name: train
num_bytes: 20007178207.559086
num_examples: 105434
- name: test
num_bytes: 2223034137.4679146
num_examples: 11715
download_size: 21504321377
dataset_size: 22230212345.027
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*