metadata
dataset_info:
features:
- name: image_path
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 20424534
num_examples: 20
download_size: 20424404
dataset_size: 20424534
configs:
- config_name: default
data_files:
- split: train
path: data/train-*