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
- [AMD] Fuse Kimi K2.5 shared-expert append into AITER grouped-topk
- [Perf] DSV4: up to -34% prefill TTFT on 840K-token prompts (B=1) via query-dim top-K reuse
- [Bug] GLM-5.3 crash on disagg decode + dp-attention + spec decode
- [AMD] Fix AITER DSA prefill with an FP8 KV cache
- [Bug] DeepSeek-V4.1-Flash + Engram: profiled SPS table (compact ragged verify) dies in CUDA-graph capture with "engram target-verify expects one equal block per request"
- [AMD] Decouple gfx942 DSA projection fusion from CUDA postprocessing
- [Diffusion] Opt in to guarded gfx942 AITER varlen strides per call
- [AMD][ROCm] Add optional device-resident ordering-edge runtime build (rocm-systems#11212)
- Fix AscendTPDispatcher downcasting FP32 routing weights (#39351)
- [NPU] Use npu.compressor for DeepSeek-V4 arch35
- Docs
- Python not yet supported