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
- [Feat] spec: Support EAGLE3 under pipeline parallelism by relaying au…
- [Fix] deepseek: Import kernel_api_logging from sglang.kernels in the AWQ dequantize helper
- [Perf] [AMD] deterministic: Faster Triton MoE config above 64 tokens on gfx950
- [Perf] diffusion: skip unread text encoder layers in FLUX.2 (~22% faster encode)
- [Perf] [AMD] deterministic: Use a fixed-order two-stage all-reduce for large inputs
- [Bug] ValueError: The checkpoint you are trying to load has model type `qwen4_exp` but Transformers does not recognize this architecture.
- [Perf] [AMD] unquant: Pad gfx11 linear and lm_head weight rows to avoid cache aliasing
- [Perf] diffusion: encode only the real Wan UMT5 tokens (5-10x faster text encode)
- [Perf] Qwen3.8: Allow breakable prefill CUDA graphs for image batches
- [Perf] mrope: Upload M-RoPE decode deltas without blocking the host
- Docs
- Python not yet supported