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
- [Refactor] multimodal: [MM refactor 5/6] Add a token-space strategy for Qwen VL models
- [Refactor] multimodal: [MM refactor 6/6] Enable token-space processing at startup with parity tests
- [Feature] [RFC] SGLang Engine Initialization Snapshots: Reusing Initialized State for Faster Startup
- [Bugfix] Preserve routed expert history across KV retraction
- support prefill-complete on npu
- [CP] Support GLM-5.3-Flash colocated prefill CP and decode attention TP with DFlash2/MTP
- [Kernel] Add H800 W8A8 block-FP8 GEMM tuning configs
- [Perf] hicache: Parallel THP-backed host pool allocation (opt-in)
- [Bug] qwen_vl: literal "<image>" in message text aborts requests with 500 "Mismatch: More IMAGE tokens found than corresponding data provided"
- [Spec] Capture the prefill CUDA graph for DeepSeek-style NextN drafts
- Docs
- Python not yet supported