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
- killing straggler daemons leaves stale files. Unlink stale files.
- adaptive spec cuda graph buffer reuse
- [PoC] Split hybrid SWA allocator into `full`/`swa` sides and rename `*TokenToKVPoolAllocator` to `*KVAllocator`
- [NVFP4 MoE Marlin fallback] w13/w2_blockscale_swizzled dead copies waste ~3.5 GiB per GPU (SM80–SM90x)
- [Feature] Enhancing gRPC control plane and operability
- Add OpAuto runtime policy for kernel backend selection
- [AMD] Enable tc_piecewise prefill CUDA graphs for DeepSeek-V4 on MI355X
- [AMD][DSv4] Replace fixed FP4 indexer logits slab with adaptive workspace
- feat(server): expose HiCache capacity in server info
- fix(grpc): enforce auth on SMG sidecar profiling endpoints
- Docs
- Python not yet supported