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
- [Attn] Make FlashInfer un-selectable for mixed-causal models
- [Bug]: DiffusionGemma: tensor shape mismatch in post_norm under concurrent decode (inputs_embeds vs silently-clamped sc_embeds)
- [XPU] [bugfix] Unify device visibility handling for XPU and align device ID selection
- [Bug]: minimax_m2 / glm47_moe / deepseek_v4 arg converters drop the last parameter when the model omits the closing tag (same class as #57699)
- [Feature]: [WideEP][CPU]: Add a CPU-only Wide Expert Parallelism well-lit path
- [Bugfix] Prevent stale collective RPC response reuse
- [Bug]: DecodeBenchConnector repeats KV cache fills for shared state tensors at large batch sizes
- [Bugfix][ROCm] Avoid RowWise FP8 scaled MM on RDNA4
- fix(eplb): async EPLB worker stall during idle periods and rank divergence (#57883)
- [Spec Decode] Enable GLM-5.2 DSpark under pipeline parallelism
- Docs
- Python not yet supported