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
- [Perf] [AMD] kimi-k3: Run the Quark FP8 KDA in-proj and MLA absorb on FP8 kernels
- [Refactor] [NPU] quantization: :recycle: Share online/offline W4A8 execution
- [Fix] [NPU] qwen3.5: :bug: Support ModelSlim mixed gate/up precision
- [Feat] [NPU] quantization: :sparkles: Add MXFP capability preflight
- [Fix] mem_cache: Reject underflowing component size deltas in the Python tree core
- [Fix] mem_cache: Fit automatic hybrid speculative state caches to available memory
- [Feat] tokenizer: add an opt-in toktier backend (exact per-session repair, chunk-parallel CPU encode)
- [Perf] DSA: Skip the unused token-level paged MQA schedule under KPool
- [Feat] diffusion/disaggregation: Add a mock transfer engine and XPU device support
- [Feat] tokenizer: await toktier's worker pool in TokenizerManager._tokenize_texts
- Docs
- Python not yet supported