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
- [Bugfix][Security] Complete input-bound hardening left incomplete by PR #51447
- docs: note flashinfer-cubin upgrade path for pip users
- [Feature]: [ROCm] Add gfx1030 (RDNA2) to Dockerfile archs + proper gate
- [Bugfix][Security] Disable ambient NETRC credentials on media fetches
- [Bug]: AITER topK meta buffer ignores pcp_size — every PCP start fails with fused shared experts on ROCm
- [Docs] Refresh Hybrid KV-cache Manager design doc
- [Perf][MoE] Scale the default MoE N tile with batch size
- [Perf][MoE] Stop demoting FlashInfer CUTLASS MoE on Hopper
- [Spec Decode] Make the dynamic SD schedule observable
- [Bug]: On ROCm, DCP and non-FP8 KV dtypes are unreachable, but the errors suggest fixes that don't exist there
- Docs
- Python not yet supported