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 supported61 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Perf][Spec Decode][GLM-5.3] Specialized GLM-5.3 MTP drafter, with a post-forward CUDA graph
- [KV Events] Coalesce consecutive BlockRemoved events
- [ROCm][Perf][DSv4] gfx1250 W4A4 MoE: fused silu + fused quant + preshuffled weights
- [ROCm][Perf][DSv4] Enable custom allreduce on gfx1250
- [Bugfix][Structured Output] Allow disable_any_whitespace with the auto backend
- [Bugfix] Fix illegal memory access when MTP/DFlash/DSpark drafts run past their max_position_embeddings
- [Bugfix][Muse] Flush header-less output as content at stream end
- [Refactor][KV Connector] Share the Prometheus registration between NIXL and Mooncake
- [RFC]: Reduce CPU overhead in Model Runner V2 for long-context P/D serving
- [Bug]: [ROCm] AutoAWQ checkpoints never use the RDNA W4A16 kernel; AWQ decodes 2.5x slower than BF16 on gfx1151
- Docs
- Python not yet supported