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
- Warm up real serving paths before the first user request
- Deprecate --num-continuous-decode-steps
- custom_mask grows unbounded via torch.cat in EAGLE/DFlash speculative decoding
- fix: exclude non-persistent buffers from weight checks
- [NPU] Support Qwen3.8-Flash-Next with graph replay, MTP, and Triton kernels
- Fix Kimi-K3 required tool call close-token loops
- [Bugfix] Fix streaming session handling with retained slots
- [Bug] Eager hybrid target verify fails when TP padding breaks fixed-width request groups
- [Fix] Match KPool draft-extend writes to precomputed plans
- [Bug] DP-attention scheduler crashes (NoneType len) when attn_tp_size>1 and attn_cp_size>1: request-broadcast source rank has work_reqs=None
- Docs
- Python not yet supported