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
- Improve multimodal processor import diagnostics
- [Bugfix] Fail clearly when DCP decode hits the LSE-less aiter path
- Fix DeepEP v2 ElasticBuffer construction inside CUDA graph capture
- MiMoV2: do not force flashinfer_trtllm when an EP a2a backend is active
- Support --moe-a2a-backend deepep_v2 for MiMoV2
- [Bugfix] Defer the QSA indexer import in the EAGLE draft worker
- [AMD] Avoid wide page-table expansion in chunked DSA prefill
- [BugFix] Fix speculative auxiliary capture after fused all-reduce
- Cuda Graph Serialization
- [Bug] SGLANG_GRAPH_BATCH_CAPTURE mislabels capture traces and crashes with IndexError when a bs bucket captures multiple graphs
- Docs
- Python not yet supported