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.
Open diffusion language model for code generation — releasing pretraining, evaluation, inference, and checkpoints.
| Date | Stars |
|---|---|
| 2026-07-31 | 644 |
| 2026-08-06 | 645 |
| 2026-08-14 | 646 |
| 2026-08-18 | 647 |
| 2026-08-20 | 646 |
| 2026-08-22 | 647 |
| 2026-08-23 | 647 |
| 2026-08-24 | 648 |
| 2026-08-25 | 649 |
| 2026-08-28 | 649 |
| 2026-08-30 | 649 |
| 2026-09-01 | 650 |
| 2026-09-04 | 652 |
| 2026-09-09 | 653 |
| 2026-09-13 | 654 |
| 2026-09-16 | 655 |
| 2026-09-17 | 655 |
| 2026-09-18 | 656 |
| 2026-09-27 | 657 |
| 2026-09-30 | 658 |
| 2026-10-01 | 659 |
| 2026-10-02 | 660 |
| 2026-10-13 | 661 |
| 2026-10-16 | 662 |
| 2026-10-27 | 661 |
| 2026-10-29 | 662 |
| 2026-10-31 | 664 |
| 2026-11-04 | 664 |
Today
— stars today
This week
+3 stars this week
This month
+4 stars this month
Momentum
0.0
growth rate 0.45%/day
# 🔥 Open-dLLM: Open Diffusion Large Language Models
🌍 Languages: [English](README.md) | [中文](README_cn.md) | [日本語](README_ja.md)
👉 TL;DR: **Open-dLLM** is the most open release of a diffusion-based large language model to date —
including **pretraining, evaluation, inference, and checkpoints**.
#### Representation Alignment
Open-dLLM supports **representation alignment** for adapting autoregressive LMs into diffusion LMs with 4x speedup. This feature is based on our recent paper, [**Don’t Retrain—Align: Adapting Autoregressive LMs to Diffusion LMs via Representation Alignment**](https://arxiv.org/pdf/2605.06885). Check out [Representation Alignment Tutorial](docs/representation_alignment.md).
<p align="center">
<a href="https://github.com/pengzhangzhi/Open-dLLM">
<img src="https://cdn.jsdelivr.net/gh/devicons/devicon/icons/github/github-original.svg" width="40" alt="GitHub"/>
</a>
<a href="https://oval-shell-31c.notion.site/Open-Diffusion-Large-Language-Model-25e03bf6136480b7a4ebe3d53be9f68a?pvs=74">
<img src="https://upload.wikimedia.org/wikipedia/commons/e/e9/Notion-logo.svg" width="40" alt="Notion"/>
</a>
<a href="https://huggingface.co/fredzzp/open-dcoder-0.5B">
<img src="https://huggingface.co/front/assets/huggingface_logo-noborder.svg" width="40" alt="Hugging Face"/>
</a>
</p>
<p align="center">
<b>💻 Code</b> | <b>📖 Blog</b> | <b>🤗 Model</b>
</p>
## 🎥 Demo
<p align="center">
<img src="https://github.com/pengzhangzhi/dLLM-training/blob/main/assets/quick-sort-demo.gif"
alt="Quick Sort Demo" width="600"/>
</p>
<p align="center"><i>QuickSort generation using Open-dCoder (0.5B)</i></p>
<p align="center">
<a href="https://youtu.be/d8WrmvUhO9g">
<img src="https://img.shields.io/badge/YouTube-Video-red?logo=youtube" alt="YouTube link"/>
</a>
<a href="https://www.bilibili.com/video/BV1ZveSz3E1J/">
<img src="https://img.shields.io/badge/Bilibili-视频-blue?logo=bilibili" alt="Bilibili link"/>
</a>
</p>
---
## ✨ Highlights
- 🏋️ **Pretraining pipeline + open datasets**
- ⚡ **Inference scripts** — easy sampling & generation
- 📊 **Evaluation suite** — HumanEval, MBPP, Infilling (lm-eval-harness + custom metrics)
- 📦 **Weights + checkpoints** on Hugging Face
- 🤝 **Transparent configs** for full reproducibility
---
## Why Open-dLLM?
Most diffusion LLM repos (e.g., LLaDA, Dream) only release **inference scripts + weights**, which limits reproducibility.
**Open-dLLM** is the first to open-source the **entire stack** for diffusion LLMs.
👉 With Open-dLLM, you can go from **raw data → training → checkpoints → evaluation → inference**, all in one repo.
---
## 🔎 Transparency Comparison of Diffusion LLM Releases
| Project | Data | Training Code | Inference | Evaluation | Weights |
|-------------------------------------------------------------------------|:---:|:-------------:|:---------:|:----------:|:-------:|
| **Open-dLLM / Open-dCoder (ours)** | ✅ | ✅ | ✅ | ✅ | ✅ |
| [LLaDA](https://github.com/ML-GSAI/LLaDA) | ❌ | ❌ | ✅ | ⚠️ Limited | ✅ |
| [Dream](https://github.com/HKUNLP/Dream) | ❌ | ❌ | ✅ | ⚠️ Limited | ✅ |
| [Gemini-Diffusion](https://deepmind.google/models/gemini-diffusion/) | ❌ | ❌ | ❌ | ❌ | ❌ (API only) |
| [Seed Diffusion](https://seed.bytedance.com/seed_diffusion) | ❌ | ❌ | ❌ | ❌ | ❌ (API only) |
| [Mercury](https://www.inceptionlabs.ai/introducing-mercury-our-general-chat-model) | ❌ | ❌ | ❌ | ❌ | ❌ (API only) |
✅ = fully available · ❌ = not provided · ⚠️ = partial/limited
---
## ⚙️ InstalExcerpt of 12,477 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:01ed0c8f5aebdd69, topic:large-language-models
matched fp:01ed0c8f5aebdd69, topic:diffusion-models
