metadata
dataset_info:
features:
- name: image
dtype: image
- name: mask
dtype: image
- name: masked_image
dtype: image
- name: edit_prompt
sequence: string
- name: src_info
dtype: string
splits:
- name: train
num_bytes: 825562842
num_examples: 999
download_size: 825464769
dataset_size: 825562842
configs:
- config_name: default
data_files:
- split: train
path: data/train-*