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
- [Feat] mamba: Support strided destination conv-state pools in the conv-window scatter
- [Refactor] [NPU] dsv4: consume registered sgl-kernel-npu Transformer APIs
- [Feat] qwen4_exp: Support the MXFP8 PLE table of compressed-tensors checkpoints
- [Feat] [XPU] graph: add context-aware support for XPU graph
- [Perf] [AMD] dsa: In-tree gfx950 paged MQA logits for decode / MTP-verify rows
- [Fix] eplb: Skip draft recording and reuse NPU TP groups
- [Model] Add Qwen3.5 sequence classification support
- [Perf] DFlash: Use CUDA graphs for draft and idle verify under DP attention
- [Feat] multimodal: Share CPU preprocessing artifacts across tokenizer workers
- [Feat] decisions: `--decision-config`, per-type default temperatures for ordinary chat models
- Docs
- Python not yet supported