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
- [Feat] model: Add Flint (RecursiveSI/flint) support
- [Feat] pd: Defer single-layer NEXTN draft prefill to decode workers
- [Fix] diffusion: float32 Plucker rays for bf16 camera poses
- [wip][Feature] add eagle4 spec decode
- [Perf] QSA: Remove the D2D copy in prefill generated kv gather
- [Feat] glm5: Expose per-layer MLA key buffer and use_fused_topk in binding
- [Feat] disaggregation: Support shared Mooncake staging buffers for PD role switching
- [Perf] hicache: Add opt-in async ready-count sync
- [Fix] diffusion: LoRA merge cache key identity and eviction
- [Fix] diffusion: LoRA adapter-state and merge-transition fixes
- Docs
- Python not yet supported