metadata
dataset_info:
features:
- name: image
dtype: image
- name: guide
dtype: image
- name: text
dtype: string
- name: mask_image
dtype: image
splits:
- name: train
num_bytes: 95968039
num_examples: 87
download_size: 88695996
dataset_size: 95968039
configs:
- config_name: default
data_files:
- split: train
path: data/train-*