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] scheduler: Keep beam group leaders out of priority preemption
- [Fix] dflash: Route a from-disk checkpoint to the model that owns it
- [Chore] deps: Bump FlashInfer to 0.7.2.post1
- [Fix] [AMD] aiter: Fix garbage decode output with unified attention at page size 1 under CUDA graph
- [Fix] mnnvl: Match FlashInfer's arch-specific CuTe DSL AR profiles
- [Feat] [AMD] mori: MORI EPv2 AsyncLL dispatch/combine, with shared experts overlapped under SBO
- [Fix] memory: Add per-rank device-memory accounting and an opt-in all-rank allocator-cache release (#46831)
- [Fix] hicache: Look up L3 below a device-only anchor under write-back
- [Fix] anthropic: Preserve cache_salt for Messages requests
- [CI] test: Defer serving benchmark imports
- Docs
- Python not yet supported