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] [NVIDIA] deterministic: fix MLA backend selection on SM100
- [Perf] qwen4: pin the offloaded PLE table at its exact size
- [Feat] qwen4: --ple-offload-token-embedding keeps embed_tokens in host memory
- [Fix] routed-experts: return no tokens in the abort echo of a request whose routing is gone
- [Feat] routed-experts: routed_experts_start_len=-1 returns one row per generated token
- [Fix] hicache: carry routed-experts rows through the host tier
- [Perf] [NVIDIA] moe: Fused NVFP4 SwiGLU MoE kernel for SM120
- [Feat] scheduler: log host-device syncs by stage and call site
- [Feat] hicache: debug check that load-backs restore the backed-up rows
- [Feat] weights: let a custom weight loader reject a tensor update cleanly
- Docs
- Python not yet supported