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
- [Feat] disaggregation: Support DSA cache layer split with the NIXL transfer backend
- [Refactor] scheduler: Give the PD queues and preemption the retirer explicitly
- [Chore] codeowners: paged-experts
- [Feat] runway praxis 1 support
- [Feat] http_server: py-spy dump every sglang process when /health fails (opt-in)
- [Perf] diffusion: Make conditioning cache eviction O(1) and raise its entry limit to 4096
- [Fix] scheduler: pause generation in abort mode to ensure idle release
- [Fix] rust-server: Name the type and retryability of streamed /generate errors
- [Feat] disaggregation: Make the PD top-logprobs metadata width configurable
- [Feat] [AMD] diffusion: Run MiniMax-H3-Image on ROCm and add MI300X/MI325X/MI350X/MI355X cookbook recipes
- Docs
- Python not yet supported