metadata
dataset_info:
features:
- name: train
struct:
- name: caption
dtype: string
- name: image_url
dtype: string
splits:
- name: train
num_bytes: 8824789
num_examples: 13088
download_size: 2812289
dataset_size: 8824789
configs:
- config_name: default
data_files:
- split: train
path: data/train-*