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
- [RFC] Expose allocator-issued page generations to external linkers
- [Perf] attention: avoid concatenating MLA queries for zigzag CP
- [Feature] hicache: support DCP with EAGLE/MTP speculative decoding
- [Fix] openai: Reject reserved special tokens in DeepSeek V4.1 requests
- [Doc] speculative: Add LongSpark
- [Perf] kda: Remove unused d/k rings from speculative ReplaySSM
- [RFC][DSpark] Online latency correction for confidence-scheduled verification
- [Bug] DSpark spec decode: FlashInfer softmax outputs q > 1.0, triggering unconditional rejection in the accept kernel
- [Fix] dspark: clamp FlashInfer softmax output to <= 1.0 for spec accept
- [Fix] [NVIDIA] moe: Pick a supported runner for auto-detected ModelOpt NVFP4 MoE on SM110/SM120
- Docs
- Python not yet supported