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
- [Fix] Refuse state-dict checkpoints that would silently drop MXFP4 triton_kernels expert weights
- [Bug] --strip-thinking-cache + retraction: release_kv_cache frees KV slots the radix tree still owns (double free)
- Leave a completion with no suffix as the prompt
- ci: diffusion-only option for the runner utilization report
- Support page_first_kv_split in MLA HiCache L3 flat page I/O
- [Quantization] Serve NVFP4-packed QAD checkpoints (PLE NVFP4 + modelopt_mixed MXFP8)
- sglang-mm: add InternVL server-pipeline processor
- [DSV4.1] Fuse source prefill Top-K and block amax
- [DeepSeek-V4.1] Extend norm and MoE fusion batch limits
- [Session] Keep session resources alive until requests finish
- Docs
- Python not yet supported