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
- [Bug] Kimi-K3 DCP a2a produces garbled decode output on 2-node GB300; NCCL_GRAPH_REGISTER=0 restores smoke-test output
- [Roadmap] Beam Development & Optimization
- [Fix] qwen-vl: Enable ModelOpt NVFP4 vision quantization.
- [Tracking] GLM-5.3-Flash performance optimization on Hopper (SM90)
- [Perf] deepseek-v4: Score a request's verify rows as one request in the C4 indexer
- [Fix] [NVIDIA] deepseek_v4: Build the DeepGEMM schedule for the low-ratio prefill-graph indexer on SM12x
- [CI] [MLU] slash-command: Register MLU with shared opt-in CI tiers
- [Fix] attention: Mask empty splits in verify_splitkv stage2 combine
- [Fix] [CPU] qk_norm: Fall back for BF16 without AVX-512 BF16
- [Feat] diffusion/disaggregation: Add declarative stage deployments
- Docs
- Python not yet supported