metadata
dataset_info:
features:
- name: subject_name
dtype: string
- name: subject_images
dtype: image
- name: source_images
dtype: image
- name: identifier
dtype: string
- name: source_prompt
dtype: string
- name: target_prompt
dtype: string
- name: add_bounding_box
sequence: int64
- name: task_type
dtype: string
splits:
- name: train
num_bytes: 189933687
num_examples: 600
download_size: 98482303
dataset_size: 189933687
configs:
- config_name: default
data_files:
- split: train
path: data/train-*