D
dLLM-RL
Gen-Verse/dLLM-RL
[ICLR 2026] Official code for TraceRL: Revolutionizing post-training for Diffusion LLMs, powering the SOTA TraDo series.
★517stars
Python
Apache-2.0
Updated: 1d ago
📋 Project at a Glance
Tap to expand
What's this?A open-source Models project, built with Python, focusing on code-generation and diffusion-language-models
Who made it?Maintained by Gen-Verse team, 517⭐ on GitHub, #1781 out of 3201 in Models
Why does it exist?In the Models space, code-generation workflows faced efficiency bottlenecks. dLLM-RL was built by Gen-Verse to address these diffusion-language-models challenges.
What can it do?Key use cases: large-language-models, llm-reasoning, mathmatical-reasoning
How to install with AI?Use an AI coding assistant (Claude Code, Cursor, Copilot) to automatically set up pip dependencies and virtual env. Follow the README — the AI handles the rest.
🔗 github.com/Gen-Verse/dLLM-RL | 官网 https://arxiv.org/abs/2509.06949
🔗 github.com/Gen-Verse/dLLM-RL | 官网 https://arxiv.org/abs/2509.06949
Topics
code-generationdiffusion-language-modelslarge-language-modelsllm-reasoningmathmatical-reasoningreinforcement-learning-algorithmsrlhf