metadata
dataset_info:
features:
- name: image
dtype: image
- name: path
dtype: string
- name: caption
dtype: string
splits:
- name: train
num_bytes: 15031210
num_examples: 22
download_size: 15032532
dataset_size: 15031210
configs:
- config_name: default
data_files:
- split: train
path: data/train-*