metadata
dataset_info:
features:
- name: edit_prompt
dtype: string
- name: original_image
dtype: image
- name: edited_image
dtype: image
splits:
- name: train
num_bytes: 68668607470.811
num_examples: 38357
download_size: 70759733399
dataset_size: 68668607470.811
configs:
- config_name: default
data_files:
- split: train
path: data/train-*