metadata
dataset_info:
features:
- name: image
dtype: binary
- name: query
dtype: string
- name: label
dtype: string
- name: imgname
dtype: string
- name: sample_id
dtype: int64
splits:
- name: train
num_bytes: 10459046387
num_examples: 300000
download_size: 9837329842
dataset_size: 10459046387
configs:
- config_name: default
data_files:
- split: train
path: data/train-*