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
- [AMD][DSv4/DSpark]: Fuse the draft block's attention-metadata prologue to graph mode
- [Bug] --preferred-sampling-params is inert on /v1/chat/completions (re-report of #21816; the five get_param keys are affected too)
- [NPU] Decode context parallelism (DCP) and prefill query sharding for DSA models
- docs: deploy SGLang on Nebius Serverless Endpoints
- perf: serialize completion stream chunks directly
- [AMD] [GLM5.3] Tune GLM-5.3-Flash KDA prefill kernels on gfx950
- [ROCm][DSpark] Fix two startup crashes when loading DeepSeek-V4-Pro-D…
- [ROCm][DSpark] Make model loading barrier timeout configurable, raise…
- [Router] Tell an upstream socket give-up apart from an exhausted request budget
- [Ollama] Map options.seed to the sampling_seed field
- Docs
- Python not yet supported