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 supported62 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Performance] Memoize waiting prefix cache hits across retries
- [Perf][DSv4.1] DSpark draft on the specialized decode
- [ROCm][Perf] Opt-in flag for the act+MXFP4 quant fusion
- [Bugfix][Attention] Use a 32-wide KV tile for FP8 queries in grouped DiffKV launches
- [ROCm][GLM-5.3-Flash] Enable KDA prefill checkpoints on the Triton backend
- [Feature][KV Connector][Mooncake] Add MooncakePullConnector for decode-side reads
- [Perf][MRv2] Add static split selection for token logprobs
- [Performance]: Monolithic Triton top-k kernel falls back to full-vocab search on real logits (outlier gather misses)
- [Bugfix][Multimodal] Reject prefill caps smaller than an intact item
- [Feature][AuxOutput] Support routed experts with P/D disaggregation
- Docs
- Python not yet supported