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
- [Bug]: AsyncScheduler num_output_placeholders underflow with chunked prefill + concurrency (no spec decode, no preemption) — regression from 0.24.0
- [Feature]: GLM 5.3 Performance Optimization
- [Core][KV Connector] Share in-flight external-prefix KV loads
- [Bug]: FlashInfer autotune config cache hits only on rank 0, deadlocking the engine launch
- [Bug][ROCm] Engine stops making progress inside an aiter fused MoE step on gfx942, then dies with no error (GLM-5.3, TP8, v0.29.0)
- [Bug]: AriaForConditionalGeneration.load_weights discards loaded parameter set, making weight tracking ineffective
- fix: resolve issue 55927
- docs: document Qwen system message ordering for Codex
- [BugFix][XPU] Skip Qwen4Exp registry import test on XPU
- [Feature][Sampling] Add native GPU output no-repeat n-gram masking
- Docs
- Python not yet supported