sglang
https://github.com/sgl-project/sglang
Python
SGLang is a high-performance serving framework for large language models and multimodal models.
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 supported1 Subscribers
Add a CodeTriage badge to sglang
Help out
- Issues
- [Fix] hicache: Separate storage keys for incompatible host layouts
- [Fix] deepgemm: adapt mhc calls to the {1,m,n} output contract (#43743)
- [Fix] [NPU] CI: Update NPU failure analyzer coverage
- [Fix] responses: fix reasoning.effort in non-streaming responses and max_output_tokens=1 validation
- [Feat] [NPU] Diffusion: support online W8A8 INT8 quantization
- [Feat] MiniMax-H3: add an opt-in four-step SMD sampling preset
- [Bug] `pause_generation(mode="retract")` while a request's grammar is compiling, then `release_memory_occupation` with an AssertionError kills the scheduler
- [Bug] When client disconnect with `pause_generation(mode="abort")` and `release_memory_occupation` kills the scheduler
- [Bug] Grayscale JPEG's processing failes on normalization step
- [Fix] nemotron-h: Allow explicit CuTe DSL W4A16 MoE
- Docs
- Python not yet supported