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
- [BugFix] Prefer per-spec cache_dtype_str when reshaping KV cache
- [Frontend][Core][Spec Decode] Per-request acceptance stats in OpenAI API responses
- [RFC]: Unify the Weight Loading Lifecycle Across Initial Load and Reload
- [Feature]: FlashInfer-CUTLASS NVFP4 MoE lacks SWIGLUSTEP activation (blocks Step-3.7-Flash-NVFP4 at TP=8)
- [Bugfix][Model] Fix MiniMax-M3 NVFP4 inference correctness
- [Bug] DeepSeek-V4 tool-call parser leaks raw DSML into content when the model omits the `<|DSML|tool_calls>` START token (long context)
- [Spec Decode] Fix speculators-format DSpark loading; mirror repetition penalty on DSpark draft logits
- [Benchmark] KV Cache Offload Benchmark — Block Copy Performance
- [ROCm] Support MiniMax-M3 NVFP4 SwiGLU-OAI
- Fix special-token leakage into content when tools/tool_choice is omitted on a later agent-loop turn
- Docs
- Python not yet supported