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] Abort held decode rebootstrap requests during pause
- [cake_kernels] kvcache / sampling / quantization adapters + shared helpers
- [cake_kernels] GEMM adapters (grouped FP8, NVFP4, Kimi-K3 FP8 projection, BF16 BMM, router GEMM, SVDQuant, DeepGEMM)
- [cake_kernels] MoE adapters (routing, fused experts, LatentMoE, mega-gate, A2A-adjacent ops)
- [cake_kernels] attention adapters (dense FMHA, MLA, sparse / DSV4, misc)
- [cake_kernels] communication adapters (MoE A2A, Kimi-K3 TP12 tail, fused norm + combine)
- [cake_kernels] MiniMax-H3 diffusion and multimodal adapters
- [cake_kernels] KDA / GDN linear-attention and Mamba adapters
- [GDN] Add --enable-linear-lossless-verify: bit-exact tuple replay for spec verify
- feat(rocm): pass a static kv_len_hint into mla_gluon decode
- Docs
- Python not yet supported