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
- [Perf] dsv41: Write large eager mHC output directly into MegaMoE buffers
- [Feat] cache: report measured prompt writes across serving APIs
- [Feat] kimi-k3: bound rank-local vision encoding batches
- [Fix] anthropic: reject unsupported nested tool-result blocks
- [Fix] diffusion/disagg: Recompute automatic topology per role
- [Feat] [NVIDIA] modelopt: Load NVFP4 experts with CSF-compressed scales
- [Perf] diffusion: Wan-Animate-2 decode clamp, LTX-2 Bongmath closed form, rollout log-prob/CPU-copy, validators NaN gate, in-place decode normalize
- [Fix] lora: Skip LoRA release for requests rejected before adapter acquire
- [Fix] api: reject malformed tool arguments in Responses and Messages
- [Perf] mooncake: Cut the host time of direct linker layer-wise loads
- Docs
- Python not yet supported