metadata
dataset_info:
features:
- name: image
dtype: image
- name: id
dtype: int64
- name: image_url
dtype: string
- name: embeddings
sequence: float32
splits:
- name: train
num_bytes: 8706048
num_examples: 500
download_size: 9052322
dataset_size: 8706048
configs:
- config_name: default
data_files:
- split: train
path: data/train-*