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
- [MoE] Default-on CUDA-graph pad-token mask for fused_moe_triton path
- [feat] DeepSeek V4 support W4A8(MXFP4FP8) on hopper
- [4/14] Route layernorm through batch-invariant RMSNorm
- [6/14] Cover TP-invariant math operations
- [8/14] Preserve DP attention shape and dtype under contract
- [11/14] Cover contract and Qwen3 MoE policy helpers
- [12/14] Cover DP and communicator true-on-policy helpers
- [13/14] Cover reduction and legacy namespace helpers
- [Gemma4] perf: batch vision encoder and embed_vision calls
- feat(sampling): add MinerUNoRepeatNGramLogitProcessor (no_repeat_ngram_size for MinerU2.5)
- Docs
- Python not yet supported