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
- [Feature] [Intel] weight-cache: Support the weight cache daemon on Intel XPU
- [Feat] [AMD] qwen4_exp: Use the packed PLE prefill short conv on ROCm
- [Bug] dsv4: Eager SWA bounded replay masks valid TP-to-EP tail rows
- [Fix] dsv4: Rebind eager replay tail and MoE shard counts
- [Fix] attention: Widen GDN chunk-state offsets to int64 before multiplying
- [Fix] compressed-tensors: Treat actorder aliases like their canonical form in WNA16 MoE
- [Fix] models: Forward router top_k to fused MoE experts
- [Fix] [AMD] kv-cache: Reject unsupported FP8 E5M2 requests
- [Feat] [NVIDIA] attention: Extend LiteTopK prefill to support LongCat, ERNIE 5.5, HY-4-family, Kimi K3.1
- [Perf] diffusion: Replay the Qwen3 text encoder from breakable CUDA graphs
- Docs
- Python not yet supported