---
license: openrail++
library_name: diffusers
dataset_info:
  features:
  - name: caption
    dtype: int64
  splits:
  - name: train
    num_bytes: 2929653589
    num_examples: 1000
  download_size: 2929757570
  dataset_size: 2929653589
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

# Margin-aware Preference Optimization for Aligning Diffusion Models without Reference

<div align="center">
<img src="assets/mapo_overview.png" width=750/>
</div><br>

We propose **MaPO**, a reference-free, sample-efficient, memory-friendly alignment technique for text-to-image diffusion models. For more details on the technique, please refer to our paper [here] (TODO). 

## Developed by

* Jiwoo Hong<sup>*</sup> (KAIST AI)
* Sayak Paul<sup>*</sup> (Hugging Face)
* Noah Lee (KAIST AI)
* Kashif Rasul (Hugging Face)
* James Thorne (KAIST AI)
* Jongheon Jeong (Korea University)

## Dataset

This dataset is *pixel art* split of **Pick-Style**, self-curated with [Stable Diffusion XL](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0). Using the context prompts (i.e., without stylistic specifications), we generate (1) cartoon style generation with stylistic prefix prompt and (2) normal generation with context prompt. Then, (1) is used as the chosen image, and (2) as the rejected image. The *chosen* field comprises pixel art style generations from SDXL, while the *rejected* field comprises the ordinary generations from SDXL.


## Citation

```bibtex
@misc{todo,
    title={Margin-aware Preference Optimization for Aligning Diffusion Models without Reference}, 
    author={Jiwoo Hong and Sayak Paul and Noah Lee and Kashif Rasuland James Thorne and Jongheon Jeong},
    year={2024},
    eprint={todo},
    archivePrefix={arXiv},
    primaryClass={cs.CV,cs.LG}
}
```