vllm-project/vllm

[Feature]: Integrate fused `kMoEFinalizeARResidualRMSNorm` from FlashInfer

Open

#40.544 aperta il 21 apr 2026

Vedi su GitHub
 (5 commenti) (1 reazione) (0 assegnatari)Python (16.816 fork)batch import
feature requesthelp wanted

Metriche repository

Star
 (80.034 star)
Metriche merge PR
 (Merge medio 9g 2h) (921 PR mergiate in 30 g)

Descrizione

🚀 The feature, motivation and pitch

Available today via FlashInfer 0.6.8, we can leverage this fused MoE Finalize + ResidualAdd + AllReduce + RMSNorm https://github.com/flashinfer-ai/flashinfer/pull/2982.

This should be added as a new torch.compile custom pass. I think moe_finalize might still be inside the wrapped fused_moe op so this might require pulling that out.

Alternatives

No response

Additional context

No response

Before submitting a new issue...

  • Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

Guida contributor