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
- [Feat] minimax: Support MiniMax-M3.1-Flash
- [Fix] [AMD] diffusion: Default to torch_sdpa and keep NCDHW VAE convs on Radeon (RDNA)
- [Feat] Comm: Gated background transfers
- [Feat] Comm: Add gated allgather in kv cache sharding
- [Perf] LoRA: improve throughput and GPU memory usage for multi-LoRA serving under PP
- [Fix] systemone: Abort scoring requests when the client disconnects and honor a caller-supplied rid
- [Feat] afd: add the attention-FFN disaggregation framework
- [Doc] developer_guide: Use sglang.benchmark.serving in the benchmark guides
- [Feat] diffusion: Native SANA-Sprint support
- [Feat] disaggregation: Support SWA hybrid models in decode radix tree
- Docs
- Python not yet supported