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 supported21 Subscribers
Add a CodeTriage badge to vllm
Help out
- Issues
- Fused MoE LoRA Kernel Optimizations
- [RFC]: online quantization user facing API
- Fail-fast guard: block MTP on GLM-4.7-FP8 to avoid CUDA illegal access
- [Bugfix][Tool Parser] Fix Qwen3 Coder parser to stream tool call arguments
- [Bug]: v0.13 CPU fails with more than 16 OMP threads
- [Bug]: Regression in v0.14.0: "No valid attention backend found" for nvidia/DeepSeek-R1-0528-NVFP4 on RTX Pro 6000 (Blackwell)
- [Bugfix][ROCm] Fix DeepSeek-R1 repetition via hybrid sampler routing
- [Feature]: `kv_cache_dtype="auto"` should resolve to the actual `kv_cache_dtype` picked by vllm, and be displayed to users
- [Bugfix] Remap FP8 scale names for Mistral3/Pixtral models
- [Bugfix][MoE] Remove tests/v1/e2e/test_spec_decode.py::test_mtp_correctness[deepseek]
- Docs
- Python not yet supported