metadata
dataset_info:
features:
- name: caption
dtype: string
- name: url
dtype: string
- name: image
dtype: image
- name: canny
dtype: image
splits:
- name: train
num_bytes: 907334380
num_examples: 10000
download_size: 191406986
dataset_size: 907334380
configs:
- config_name: default
data_files:
- split: train
path: data/train-*