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
- [PD] Decode a copy of the running rows and drop finished rows at commit
- [PP] Join and drop a slot's running rows at the ring commit
- [Test] Record the call order of every event loop on a fake scheduler
- [Scheduler] Run the flat event loops through one driver
- [PP] Run the pipeline loops through one ring driver
- [Feature] SM90 W8A16 small-M fused MoE for 128x128 block-FP8 experts
- [Perf] SM90 W8A16 small-M fused MoE for 128x128 block-FP8 experts
- [Fix] Cap transcription upload size before buffering it in host memory (#43709)
- Fix LoRA loading for ungated MoE up_proj exports
- [DSA] Split topk across CTAs for small CUDA BF16 TileLang sparse decode
- Docs
- Python not yet supported