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
- [Perf] [AMD] DeepSeek-V4.1: Opt-in SDMA exchange for the HIP indexer row shard
- [Perf] [AMD] DeepSeek-V4.1: Faster HIP dense prefill indexer (row-group scoring, |s| epilogue, tile-gated top-k)
- [Fix] [AMD] topk-v2: Fix wrong picks on split-launch short rows
- [Perf] attention: Parallelize MLA chunked-prefix KV index construction
- [Feature] Add bitsandbytes support for Qwen3-VL / Qwen3-VL-MoE
- [Fix] hicache: Wait for overlapping write-through acknowledgements
- [Perf] [NVIDIA] bf16-gemm: SM107 split-K and cute_ext tactics
- [Perf] [XPU] fp8: Run FP8 dense layers weight-only (W8A16) through sgl_kernel.fp8_linear
- [Fix] kimi-k3: Rebuild derived weights in place on weight updates
- [Perf] [XPU] vae-layernorm-silu: add Xpu triton path to replace torch
- Docs
- Python not yet supported