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] Add UltraQuant 4-bit KV cache backend (FlyDSL D=256).
- [PD] Expose unified memory transfer layouts
- [Sampling] Default to `top_k_first` and expose filtering order
- [Perf] Fuse compact `top_k_first` probability construction
- [Fix] Align Qwen-VL JPEG decoding and repetition penalty semantics
- feat(perf): Add compact Mamba2 speculative-state replay
- [Quantization] Opt-in bf16 path for small-M block-FP8 attention projections
- [AMD][Diffusion] Route unquantized linears through aiter tgemm on gfx1250
- Fix usage.prompt_tokens always 0 for harmony models on /v1/responses
- [Model] Add MiniCPM-V 4.7 support
- Docs
- Python not yet supported