metadata
dataset_info:
features:
- name: image
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 30000
num_examples: 1500
download_size: 1652
dataset_size: 30000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*