metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: embeddings
sequence: float32
splits:
- name: train
num_bytes: 121979613
num_examples: 833
download_size: 103390020
dataset_size: 121979613
configs:
- config_name: default
data_files:
- split: train
path: data/train-*