Datasets:
metadata
dataset_info:
features:
- name: original_image
dtype: image
- name: edit_prompt
dtype: string
- name: bokeh_image
dtype: image
splits:
- name: train
num_bytes: 676092
num_examples: 4400
download_size: 0
dataset_size: 676092
size_categories:
- 1K<n<10K
language:
- en
task_categories:
- image-to-image