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.
dLLM: Simple Diffusion Language Modeling
| Date | Stars |
|---|---|
| 2026-07-31 | 2657 |
| 2026-08-09 | 2665 |
| 2026-08-17 | 2670 |
| 2026-08-18 | 2671 |
| 2026-08-19 | 2673 |
| 2026-08-20 | 2671 |
| 2026-08-21 | 2671 |
| 2026-08-23 | 2672 |
| 2026-08-25 | 2673 |
| 2026-08-26 | 2672 |
| 2026-08-27 | 2671 |
| 2026-08-28 | 2673 |
| 2026-08-29 | 2674 |
| 2026-08-31 | 2675 |
| 2026-09-01 | 2676 |
| 2026-09-02 | 2678 |
| 2026-09-04 | 2679 |
| 2026-09-06 | 2680 |
| 2026-09-09 | 2681 |
| 2026-09-10 | 2682 |
| 2026-09-12 | 2684 |
| 2026-09-14 | 2685 |
| 2026-09-15 | 2688 |
| 2026-09-16 | 2691 |
| 2026-09-17 | 2692 |
| 2026-09-18 | 2693 |
| 2026-09-20 | 2692 |
| 2026-09-22 | 2693 |
| 2026-09-23 | 2692 |
| 2026-09-25 | 2691 |
| 2026-09-26 | 2694 |
| 2026-09-27 | 2697 |
| 2026-09-28 | 2699 |
| 2026-09-29 | 2698 |
| 2026-09-30 | 2699 |
| 2026-10-01 | 2700 |
| 2026-10-02 | 2701 |
| 2026-10-03 | 2699 |
| 2026-10-08 | 2702 |
| 2026-10-10 | 2704 |
| 2026-10-13 | 2706 |
| 2026-10-14 | 2707 |
| 2026-10-15 | 2708 |
| 2026-10-16 | 2710 |
| 2026-10-17 | 2711 |
| 2026-10-18 | 2711 |
| 2026-10-19 | 2712 |
| 2026-10-20 | 2713 |
| 2026-10-21 | 2713 |
| 2026-10-22 | 2713 |
| 2026-10-23 | 2714 |
| 2026-10-24 | 2715 |
| 2026-10-26 | 2716 |
| 2026-10-28 | 2718 |
| 2026-10-29 | 2721 |
| 2026-10-30 | 2722 |
| 2026-10-31 | 2721 |
| 2026-11-02 | 2720 |
| 2026-11-03 | 2723 |
| 2026-11-04 | 2724 |
Today
+1 stars today
This week
+6 stars this week
This month
+25 stars this month
Momentum
0.0
growth rate 0.22%/day
<h1 align="center">dLLM</h1>
<p align="center">
Simple Diffusion Language Modeling
</p>
<!-- <div align="center">
[](https://arxiv.org/abs/2602.22661)
[](https://huggingface.co/dllm-hub)
</div> -->
<p align="center">
📃 <a href="https://arxiv.org/pdf/2602.22661" target="_blank">Report</a> | 🤗 <a href="https://huggingface.co/dllm-hub" target="_blank">Models</a>
</p>
<p align="center">
<img
src="assets/logo.gif"
alt="dLLM logo">
</p>
## Overview
**dLLM** is a library that unifies the training and evaluation of **diffusion language models**, bringing transparency and reproducibility to the entire development pipeline:
- dLLM provides scalable training pipelines (based on [`transformers`](https://github.com/huggingface/transformers/blob/main/src/transformers) [Trainer](https://github.com/huggingface/transformers/blob/main/src/transformers/trainer.py)), with support for [LoRA](https://github.com/huggingface/peft), [DeepSpeed](https://github.com/deepspeedai/DeepSpeed), [FSDP](https://pytorch.org/blog/introducing-pytorch-fully-sharded-data-parallel-api/) and beyond.
- dLLM provides unified evaluation pipelines (based on [`lm-evaluation-harness`](https://github.com/EleutherAI/lm-evaluation-harness)) that abstracts away inference details and making customization simple.
- Built on these components, dLLM provides the minimal **training / inference / evaluation** recipes for open-weight models (e.g., [LLaDA](https://arxiv.org/abs/2502.09992) and [Dream](https://arxiv.org/abs/2508.15487)), and implementations of training algorithms (e.g., [MDLM](https://arxiv.org/abs/2406.07524) (masked diffusion), [BD3LM](https://arxiv.org/abs/2503.09573) (block diffusion), [Edit Flows](https://arxiv.org/abs/2506.09018) and so on).
<!-- > [!NOTE]
> This repository is primarily for educational purposes and does not aim for 100% exact reproduction of official models (which is impossible). We hope it serves as a helpful reference for the community — contributions and improvements are always welcome! -->
## News
<!-- **[2026/02]** 📄 Checkout our **[`technical report`](assets/dLLM.pdf)**! -->
**[2026/04] 🎯 [`diffu-GRPO`](https://github.com/dllm-reasoning/d1)**: We support diffu-GRPO training for masked diffusion language models, validated on [LLaDA](https://arxiv.org/abs/2502.09992) and [Tiny-A2D](https://huggingface.co/collections/dllm-hub/tiny-a2d) across five reasoning tasks (GSM8K, MATH, Countdown, Sudoku, Code). See [`examples/rl`](/examples/rl#grpo) for training instructions.
**[2026/02] ⚡[`Fast-dLLM`](https://github.com/NVlabs/Fast-dLLM)**: We support accelerated inference and evaluation of [LLaDA](https://arxiv.org/abs/2502.09992) and [Dream](https://arxiv.org/abs/2508.15487) with [Fast-dLLM](https://arxiv.org/abs/2505.22618) (cache, confidence-threshold decoding). See [`examples/fastdllm`](/examples/fastdllm) for inference / evaluation instructions.
**[2025/12] 🤗[`Tiny-A2D`](https://huggingface.co/collections/dllm-hub/tiny-a2d)**: We released a collection of **SOTA** small (0.5B/0.6B) diffusion models adapted from AR models, with fully open recipes for converting **ANY** AR model (e.g., Qwen, LLaMA, and GPT-2) into a diffusion model. See [`examples/a2d`](/examples/a2d) for training / inference / evaluation instructions.
**[2025/11] 🤗[`BERT-Chat`](https://huggingface.co/collections/dllm-hub/bert-chat)**: We released a collection of BERTs finetuned to chat with diffusion, with open recipes for turning **ANY** BERT encoder (e.g., BERT, RoBERTa, ModernBERT) into a diffusion model. See [`examples/bert`](/examples/bert) for training / inference / evaluation instructions.
## Table of Contents
- [Features](#features)
- [Setup](#setup)
- [Files](#files)
- [Training](#training)
- [Inference](#inference)
- [EvaluationExcerpt of 16,417 characters
Read on GitHub
Zhanhui Zhou · UC Berkeley
26
Lingjie Chen · University of Illinois Urbana-Champaign
6
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:d4f851a9e8054e99, topic:llm, readme:technical report
matched fp:d4f851a9e8054e99, topic:nlp