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
- [Bugfix] Fix Dots NextN post-load initialization
- [Bugfix] Retain Dots towers for adaptive encoder dispatch
- Fix --model-path required error when using --config with sglang serve
- Add Q/K Hadamard rotation for KV cache FP4 quantization to improve the accuracy.
- [Feature] Support deterministic inference for DeepSeek-V4
- [Bugfix] Reject empty input_embeds during request normalization
- [JIT Kernel] Migrate gptq_kernel from AOT to JIT
- fix(pd): make DSV4 retraction safe
- [CP][AMD] Add runtime TP/CP switching for MLA
- [Model] Follow up dots.note.omni attention routing, MTP defaults, and video sampling
- Docs
- Python not yet supported