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
- [DO NOT MERGE] [LoRA] Fix dense TP buffers, idle batches and multi-node startup
- [AMD] [GLM5] Add opt-in PTPC FP8 KDA projections on gfx950
- [Bug] Qwen3ForSequenceClassification retains random head bias when loading bias-free checkpoints
- [Perf] Consume linear NVFP4 scales in FlashInfer CUTLASS A2A MoE
- [Feature] Add Prometheus metrics for multimodal preprocessing latency
- fix: zero the default Qwen3 classification bias
- [Bug] DSA attention: `tl.constexpr` strides in `transform_index_page_table_{prefill,decode}_kernel` cause unbounded Triton recompiles and stall
- [Bugfix] Preserve PP token shards with MoE A2A
- [Feature] Korean Localization for SGLang Cookbook
- [AMD][DSA] Extend Triton sparse-MLA prefill to gfx942 (MI308X) via 2-query merged programs
- Docs
- Python not yet supported