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
- Fix/dsv4 pre-sm90 (software fp8) sparse mla omnibus
- [Perf][Model Runner V2] GDN all-mode decode-path optimizations: packed QKV, fused index prep, weight de-interleave + GEMM fold
- [Security] Validate kv_transfer_params role sub-dicts are dicts
- [Spec Decode] Enable fused draft metadata updates for sparse MLA backends under DCP
- [Bug][XPU] inductor-compiled decode pieces x speculative decode livelock both GPUs at >=32k ctx (graph capture alone clean; native stacks included)
- [Model Runner V2][DP] Add target execution contracts for idle EP ranks
- [Bugfix] Validate kv_transfer_params before NIXL recv registration
- [Model Runner V2][DP] Cache scheduler-epoched target execution contracts
- [RFC][Core] Prototype runtime scheduler sequence cap
- [Feature][Frontend] Add reasoning and content phase per-request metrics
- Docs
- Python not yet supported