metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: original_prompt
dtype: string
- name: original_image
dtype: image
- name: edit_prompt
dtype: string
- name: edited_prompt
dtype: string
- name: edited_image
dtype: image
- name: adversarial_image
dtype: image
- name: edit_adv_image
dtype: image
splits:
- name: train
num_bytes: 86919610
num_examples: 200
download_size: 86923093
dataset_size: 86919610