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
- [Feature] Enable DCP for GLM-5.2 on SM120 with MTP support
- Fix client-disconnect CancelledError crashing the entire engine
- [Rust] Scan chat media without JSON serialization
- [Rust] Build cumulative stream frames in one buffer
- [OpenAI] Share PD/DP routing fields and adapter across entrypoints
- [Spec] Budget the DFLASH draft KV pool at the attention group's width
- [Bug] --moe-runner-backend deep_gemm accepted for DSV4.1 MXFP4 experts, then fails in CUDA graph capture (layout.hpp:108, sm_121)
- [Spec] Stabilize the DP adaptive tier vote: shared EMA + min dwell
- feat(minimax): integrate VibeCUDA MSA on Blackwell
- [Bug] DeepSeek-V4 SM120 decode pads q to 64 heads for an SM90 constraint; removing the pad changes greedy output despite the kernel being bit-identical
- Docs
- Python not yet supported