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 supported37 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [BugFix] Use file:// rendezvous for single-node executors to eliminate startup port races
- [Bug]: BF16 MoE + LoRA startup crash on non-gated models (TrtLlmBf16LoRAExperts weight-shape assert)
- [MoE] Add Platform routing backend interface
- [Bug]: MTP proposer passes are forced to PIECEWISE (`llm_base_proposer` omits `uniform_decode`), so GDN runs eager Python every step (~16 ms/step CPU orchestration, GPU idle ~85%)
- Zen5 image build torch fix
- [ROCm][MLA] Fix fp8 KV cache decode on the AITER MLA backend
- [Models] Move Qwen3.5 to hardware agnostic model definitions
- [Bug]: extract_hidden_states fails with TP>1 when KV page size < hidden-state per-token cost
- [Core] Add unit tests for BlockTable slot mapping
- [BugFix] Pick the DP world-group port at bind time via the coordination store
- Docs
- Python not yet supported