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
- [Bug] CUDA illegal memory access during compact ragged-verify graph capture (c128 plan kernel race)
- [MoE] Add SharedEP, a shared-object expert-parallel backend
- fix(router): enforce lifecycle-based admission control
- [Spec][DSV4] perf: Compile draft-extend hc_head during CUDA graph capture
- fix: account for DFlash draft KV in HybridSWAPoolConfigurator cell_size
- fix(distributed): use full TP group for fused post-MoE all-reduce
- [MLX] Preserve auxiliary state before chained decode lookahead
- [Bugfix] Thread LTX-2 linear prefixes for ModelOpt NVFP4
- [Bug] Decode starved to ~1 batch per 24s under sustained chunked-prefill load (strict prefill-first scheduling, spec decode)
- [test] Add regression test for validate_deepseek_v4_cp hook ordering (#32553; fix landed via #33532)
- Docs
- Python not yet supported