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
- [ROCm][DSv4.1] Re-enable the fused combine_topk_swa_indices on gfx950
- [Bugfix][Spec Decode] Don't overwrite a declared DSpark architecture with the Qwen3.5 MTP rewrite
- [Bug]: Whisper transcription `prompt` corrupts decoding — `<|prev|>` is not a Whisper vocab token (should be `<|startofp
- [Bugfix] Include mixed prompt embedding masks in prefix cache keys
- fix(bench): use sweep-line algorithm for exact peak concurrency in bench serve
- [Perf][DSpark] lossless improve DSpark throughput by W2 matrix topk pruned
- [Bug][HiSparse] Decode engine dies with cudaErrorLaunchFailure in the host-mirror path under sustained P/D host imports
- [Bugfix][Spec Decode][Mooncake] Fix MooncakeStore cache misses from EAGLE/DSpark-drop-induced Mamba/a…
- [Bugfix][Metrics] Release LoRA tracking when streaming input closes
- [Bug]: DSML tool-call recovery (#55954) emits ghost empty-argument calls on truncation and narration, and recovers undeclared tools
- Docs
- Python not yet supported