vllm
https://github.com/vllm-project/vllm
Python
A high-throughput and memory-efficient inference and serving engine for LLMs
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported51 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [torch.compile] Remove dead torch 2.9 Inductor monkeypatches
- [ROCm] DFlash unusable for GLM-5.3-Flash: glm5next lacks SupportsEagle3, and ROCM_AITER_MLA_SPARSE has no non-causal path
- [Perf] muse_glimmer: O(1) prefilter for the reasoning-end hot path
- [Bug]: Hybrid mamba/attention page alignment inflates attention blocks to 7808 tokens — per-request KV footprint of ~23 giant pages collapses concurrency and prefix caching (GLM-5.3-Flash/Glm5Next)
- [Bugfix][Model Runner V2] Complete a forced reasoning end that extends the natural end
- [Bug]: Qwen2.5 tool parser plus openai content format fails chat requests
- [Bug]: --enable-dbo reaches an assertion-backed all2all backend validation failure
- [RFC]: Batch invariance for speculative decoding needs to cover the forward pass (M=1 vs M=k+1)
- [Bugfix] Autotune FlashInfer decode kernels before CUDA graph capture
- [Cleanup] Decouple MoE FlashInfer utils into flashinfer_moe and clean up imports
- Docs
- Python not yet supported