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 supported42 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Core][Perf] Add MessagePack prefix block hash algorithms
- multi-turn support reasoning mode
- [Bugfix][V1] Zero recycled KV cache blocks for FullAttentionSpec to fix non-deterministic output at temperature=0
- [Test] Add parallel sampling tests for all output_kinds (AsyncLLM + LLM)
- [MoE/b12x] Add cutlass_prefill_threshold hybrid dispatch to FlashInferB12xExperts
- Add opt-in Unicode tag-block sanitization middleware
- [Bugfix][DSV4] wo_a scale getattr fallback for non-Marlin kernel paths
- [Bugfix] Fix CUDA illegal memory access in Marlin MoE c_tmp on SM 12.0a (re-file of #36889)
- [Frontend] Improve response_format validation errors for OpenAI-compatible chat/completion APIs
- [Bugfix] Fix custom compiler backend resolution
- Docs
- Python not yet supported