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 supported36 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Bug]: KV connector "Throughput (MB/s)" sums overlapping transfer durations; propose reporting avg per-transfer throughput instead
- [Bug]: Gemma4 Unified image requests produce all-NaN logits after BF16-to-FP16 fallback
- SIGSEGV in torch::jit::invokeOperatorFromPython (transpose) with NVFP4 + DFlash + torch.compile on Blackwell SM120
- LMCache + vLLM: pipeline parallelism now works with MLA models (Kimi K2, GLM-4.6)
- [Bug]: Gemma4 dense NVFP4 (nvidia/Gemma-4-31B-IT-NVFP4) fails on v0.24.0 with NotImplementedError in quant_method.tie_weights (lm_head/embed_tokens weight tying)
- [Feature]: Operator opt-out for the expandable_segments KV-connector rejection (copy-based connectors have no viable config on unified memory)
- [Core] Skip redundant draft token alloc + sampling
- [BugFix] Fix `num_output_placeholders` preemption underflow
- [Feature] PrefillDelayer for better DPA scheduling
- [Perf][ROCm] Add AITER custom AG/RS
- Docs
- Python not yet supported