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
- [Bug] diffusion_num_queue_reqs does not reflect requests pending in the ZMQ transport
- [Perf] SM90 decode regression v0.5.22 â v0.5.23: +33~45% per-step at concurrency 12 (8Kâ64K), and `use_deep_select_decode` costs acceptance on Hopper
- [Fix] model loading: Support RunAI for PP Nemotron MTP drafts
- [Feat] xpu: Consolidate simple XPU enablements
- [Fix] radix-cache: keep streaming-session KV in its prefix namespace
- [Feat] hisparse: Add the HiCache backend data plane
- [Fix] openai: Reject assistant tool calls without a function name (HTTP 500 -> 400)
- [CI] [Intel] xeon: Move all PR test suites to two SPR runners
- [Feat] [NPU] DSpark: GLM-5.2 support DSpark compact mode
- [Fix] pd: Keep a transfer Success terminal when a client abort lands late
- Docs
- Python not yet supported