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 supported60 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Security] Require trust for remote prefill cache accounting
- [Bugfix][Model] DeepSeek-V4.1: keep decoder replay off when its cut cannot be replayed
- [Bugfix][V2 runner] Keep idle-DP dummy requests within max_model_len
- [XPU][Qwen4Exp] Run the QSA side-cache ops and indexer prepare on XPU
- [XPU][Qwen4Exp] Run QSA block selection on XPU
- [Core][Scheduler] Add an opt-in time-bounded prefill hold
- [ROCm][Multimodal] FP8 ViT attention: opt-in AITER asm path (head_dim padded to 128); single-writer scale file
- [ROCm] Add VLLM_ROCM_USE_AITER_QUICK_REDUCE to keep INT4 QuickReduce on the HIP kernel
- [ROCm][DBO] QuickReduce and SDMA INT4 all-reduce for pure-TP microbatches
- [Bugfix][Benchmark] Fix timeline export at rounded minute boundaries
- Docs
- Python not yet supported