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
- [Spec Decode] H-Spec (mamba_attn_hybrid) speculative decoding with NPU (Ascend) support
- Benchmark: Qwen3.8-27B FP8 on H200 with bfloat16 SSM state (24 speed rows, 4 full GSM8K scores)
- [router] Fix power_of_two with --dp-aware
- [Bugfix] Preserve PD bootstrap fields in Rust OpenAI endpoints
- [rust-renderer] Expand scalar bootstrap rooms across parallel choices
- [Fix] GLM/Qwen3: keep a <tool_call> quoted in reasoning out of streamed tool calls
- [Renderer] Fail fast on concurrent choice errors
- Support CUDA DP attention in the shared DFlash/DFlash2 path
- fix: include Rust health probe admission in timeout
- fix: skip cancelled queued renderer tokenization
- Docs
- Python not yet supported