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] hicache: Let the PP-sync channel run the pipeline's depth ahead, not one round (#45351)
- fix(gemma4): stop array parser hanging on nested arrays with brackete…
- [Feat] fp8: Write block-FP8 linear results into caller-owned outputs
- [Perf] [NVIDIA] fp8: Tiny block-FP8 GEMM with fused UE8M0 activation quantization on SM10x
- [Perf] [NVIDIA] kimi-k3: Device-prepared expert-tile routing metadata for the FlashInfer FP4 MoE
- [Fix] attention: Use draft forward width for DSpark FA backend
- [Feat] Unified L3: support mooncake storage backend
- [Bug] MiMoDetector: string "null" becomes None, invalid booleans become false, and text around tool calls is dropped
- [Feat] diffusion: support BRIA Fibo 2.0 image generation and editing
- [Fix] [AMD] MiniMax-M3.1-Flash-Quark-Mxfp4: requantize the FP8 shared expert so fusion can load
- Docs
- Python not yet supported