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] dspark: Keep the real verify accept cost under simulated acceptance
- [Fix] dspark/MiniMax-M3.1-Flash: Inject decode-step draft KV under the draft's TP placement
- [Fix] DSpark: Make the simulated-acceptance sampling skip opt-in
- [Perf] Hunyuan3D: Speed up shape export with GPU-side grid generation
- [Perf] dsv41: Use the Engram all-to-all lookup for decode and CUDA graphs
- [Fix] disaggregation: Agree on each prefill KV transfer's outcome across PP stages
- [Test] [AMD] fp8-grid: Compare the GEMM to aiter within one bf16 rounding, not bitwise
- [Feat] rust: Parse DeepSeek-V4/V4.1 output with deepseek-recipe
- [Fix] [Intel] gemma4: Bidirectional image attention on the intel_xpu backend
- [Chore] flashinfer: Use flashinfer moe-ep API instead of deprecated NVLinkOneSidedAllToAll API
- Docs
- Python not yet supported