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] Flush buffered text at stream end in GLM-4.7/4-MoE and Cohere Command R4 detectors
- [Bugfix] Make top-k/top-p filtering consistent across sampling paths
- [Cookbook Benchmark] Qwen3.8-Flash-Next NVFP4 on B200 (TP1)
- [Bug] DCP fails for GLM-5.3-NVFP4 on B300
- [AMD] Initialize Inkling safely on ROCm
- [qwen 3.8 next] Reuse QSA prefill length metadata
- [Cookbook Benchmark] Qwen3.8-Flash-Next NVFP4 on B300 (TP1)
- [diffusion][xpu] FLUX.2: enable the bit-exact Triton SwiGLU on XPU
- fix(args): capture the exact --cuda-graph-max-bs-prefill ceiling bucket
- Benchmark: DeepSeek-V4.1-Flash B200/B300 TP4/EP4 (28 speed rows, 4 full GSM8K scores)
- Docs
- Python not yet supported