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 supported59 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [XPU][Offload] Place offloaded routed experts per expert by routing rank
- [Bugfix][Core] Re-prepare model state after extensible KV cache extension (Kimi-K3 AMD specialized decode)
- [XPU][Offload] Learn the expert rankings while serving
- [XPU][Offload] Let the hot experts follow the traffic
- [XPU][MoE] Stage UVA-offloaded experts on XPU
- [Offload] Prefetch the next layer's staged experts on a side stream
- [MiniMax-M3] Experimental: replicate attention across TP ranks for slow inter-node TP
- [Model Loader] Add opt-in scoped weight-loading thread cap
- [KV Connector] Mooncake Store: keep a request re-admitted in the step that reports its preemption
- [Docs] Explain hybrid KV cache configuration metrics
- Docs
- Python not yet supported