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] dLLM: Skip denoising for FDFO prompt-only batches
- [Perf] [AMD] Qwen4-Exp/HC: Fuse decode-sized hyper-connection norm and inject into FlyDSL mix
- [Perf] [AMD] qsa: Faster decode MQA logits on ROCm (padded Q, group-strided launch)
- [Bug] Fix Qwen-Image-2.1 config resolution for underscore directory names
- [Fix] hisparse: Use host-backed logical capacity for request limits
- [Feat] platforms: Add tensor IPC extension points
- [Fix] tbo: Preserve child offsets and auxiliary hidden states
- [Bug] TBO can change deterministic prefill outputs and drop speculative auxiliary hidden states
- [Fix] [Intel] docker/xpu: Install the Intel GPU stack from the versioned intel-omix channel
- [Perf] [AMD] DeepSeek-V4.1: Token-sharded prefill attention under attention TP
- Docs
- Python not yet supported