metadata
dataset_info:
features:
- name: original_image
dtype: image
- name: edit_prompt
dtype: string
- name: new_image
dtype: image
splits:
- name: train
num_bytes: 15287553
num_examples: 550
- name: test
num_bytes: 4336319
num_examples: 151
- name: valid
num_bytes: 1661844
num_examples: 54
download_size: 21245949
dataset_size: 21285716
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*