Open recipe to turn Qwen3 into a diffusion LLM 👀👀

> Swap the causal mask for bidirectional attention
> Source model matters a lot for performance
> Block diffusion (BD3LM) >> masked diffusion (MDLM)
> Light SFT with masking

Great work from @asapzzhou with his dLLM library! pic.twitter.com/ec2tSXAUA1

— Maxime Labonne @ ICLR (@maximelabonne) 8 décembre 2025

Open recipe to turn Qwen3 into a diffusion LLM > Swap the causal mask for bidirectional attention
> Source model matters a lot for performance
> Block diffusion (BD3LM) >> masked diffusion (MDLM)
> Light SFT with masking Great work from @asapzzhou with his dLLM library!

→ View original post on X — @maximelabonne