Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
[ECCV 2026] Official Code of "Distribution Matching Distillation Meets Reinforcement Learning"
| Date | Stars |
|---|---|
| 2026-08-04 | 290 |
| 2026-08-10 | 290 |
| 2026-08-12 | 290 |
| 2026-08-13 | 290 |
| 2026-08-18 | 290 |
| 2026-08-23 | 291 |
| 2026-08-24 | 292 |
| 2026-08-31 | 293 |
| 2026-09-02 | 294 |
| 2026-09-05 | 295 |
| 2026-09-09 | 296 |
| 2026-09-12 | 297 |
| 2026-09-13 | 298 |
| 2026-09-21 | 299 |
| 2026-09-27 | 300 |
| 2026-10-01 | 301 |
| 2026-10-03 | 302 |
| 2026-10-08 | 303 |
| 2026-10-09 | 304 |
| 2026-10-10 | 305 |
| 2026-10-16 | 305 |
| 2026-10-31 | 307 |
| 2026-11-02 | 308 |
| 2026-11-03 | 309 |
| 2026-11-04 | 309 |
Today
— stars today
This week
+4 stars this week
This month
+7 stars this month
Momentum
0.0
growth rate 1.31%/day
<h1 align="center"> DMDR:<br>Distribution Matching Distillation Meets Reinforcement Learning</h1>
<div align="center">
<a href='https://arxiv.org/abs/2511.13649'><img src='https://img.shields.io/badge/ArXiv-red?logo=arxiv'></a>
<a href="https://github.com/vvvvvjdy/dmdr"><img src="https://img.shields.io/badge/GitHub(DMDR)-9E95B7?logo=github"></a>
<a href="https://github.com/Tongyi-MAI/Z-Image"><img src="https://img.shields.io/badge/GitHub(ZImage)-9E95B7?logo=github"></a>
<a href='https://huggingface.co/Tongyi-MAI/Z-Image-Turbo'><img src='https://img.shields.io/badge/HuggingFace(ZImageTurbo)-blue?logo=huggingface'></a>
</div>
---
<p align="center">
<figcaption style="text-align: center; margin-top: 10px; font-size: 0.95em;">
Images generated by Z-Image-Turbo.
</figcaption>
<img src="assets/demo_pic_z.png" alt="Result" style="width:100%;">
</p>
---
<p align="center">
<figcaption style="text-align: center; margin-top: 10px; font-size: 0.95em;">
Images generated by SD3.5 Large finetuned through DMDR with open-source data and reward model using 4 NFE.
</figcaption>
<img src="assets/demo_pic_sd35.png" alt="Result" style="width:92%;">
</p>
---
## 🌟 Inference
**8 step Z-Image-Turbo generation (distilled by Decoupled-DMD and DMDR)**
See [Z-Image](https://github.com/Tongyi-MAI/Z-Image) repo.
## 🌠 Training
We now only have access to open the training demo code of ImageNet, hoping it can help the community to understand DMDR!
Refer to [SiT (class-conditional generation on ImageNet)](train_cc/sit/README.md) for training few-step SiT diffusion model.
## 🥂 Other Research of Our Team
- **[Decoupled DMD](https://arxiv.org/abs/2511.22677)**: Rethinking how DMD works and revealing a functional decoupling
strategy with CFG Augmentation (CA) as the primary engine for few-step conversion and Distribution
Matching (DM) as the regularizer.
## 🤝🏻 Acknowledgement
This code is mainly built upon [DMD2](https://github.com/tianweiy/DMD2), [SRA](https://github.com/vvvvvjdy/SRA), [ReFL](https://github.com/zai-org/ImageReward), repositories.
Thanks for their contributions to the community.
We also sincerely thank the opensource weights from [REPA](https://github.com/sihyun-yu/REPA), [DINOv2](https://github.com/facebookresearch/dinov2) and so on.
We only use these weights and data for research purpose.
## 🌺 Citation
If you find DMDR useful, please kindly cite our paper:
```bibtex
@article{jiang2025distribution,
title={Distribution Matching Distillation Meets Reinforcement Learning},
author={Jiang, Dengyang and Liu, Dongyang and Wang, Zanyi and Wu, Qilong and Li, Liuzhuozheng and Li, Hengzhuang and Jin, Xin and Liu, David and Li, Zhen and Zhang, Bo and others},
journal={arXiv preprint arXiv:2511.13649},
year={2025}
}
```
Excerpt of 2,889 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:cc51dcece2887aef, llm:Repository title/description: "[ECCV 2026] Official Code of 'Distribution Matching Distillation Meets Reinforcement Learning'" — indicates implementation of a distillation method combining distribution matching and reinforcement learning (model distillation research).
matched fp:cc51dcece2887aef, llm:Repository title/description: "[ECCV 2026] Official Code of 'Distribution Matching Distillation Meets Reinforcement Learning'" — indicates implementation of a distillation method combining distribution matching and reinforcement learning (model distillation research).
matched fp:cc51dcece2887aef, llm:Repository title/description: "[ECCV 2026] Official Code of 'Distribution Matching Distillation Meets Reinforcement Learning'" — indicates implementation of a distillation method combining distribution matching and reinforcement learning (model distillation research).