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
- [Fix] function_call: Preserve MiniCPM5 boundaries inside CDATA
- [Perf] HiSparse: Reduce redundant DSA index-K memory to increase decode capacity
- [Fix] nemotron: Accept transformers 5 layer types in NemotronHConfig
- [Fix] [XPU] server_args: Disable decode graph on XPU when tp_size > 1
- stop_regex: a non-literal match is left in the returned text, and no_stop_trim has no effect for regex stops
- [Refactor] pd: Extract MoriTransport behind the common Transport interface
- [Fix] stop_regex: Record the matched span, not the pattern source
- [Perf] diffusion: Skip the cross-request cache for Qwen3-VL vision features
- [Perf] diffusion: Replay Qwen-Image-2.1's text-only encoder from CUDA graphs
- [Perf] [NVIDIA] attention: Add opt-in LiteTopK prefill indexer top-k for DSA (FP8) and DeepSeek-V4.1 (FP4) on SM100
- Docs
- Python not yet supported