metadata
dataset_info:
features:
- name: image
dtype: image
- name: mask
dtype: image
- name: masked_image
dtype: image
- name: edit_prompt
sequence: string
splits:
- name: train
num_bytes: 22694698
num_examples: 30
download_size: 22696292
dataset_size: 22694698
configs:
- config_name: default
data_files:
- split: train
path: data/train-*