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] glm5.3flash: Allow PD prefill CP with native MTP or without speculative decoding
- [Fix] dspark: Forward target capture outputs from verification
- [Perf] [NVIDIA] dsa: Run FlashInfer's fused SM120 indexer for DSA top-k
- [Perf] DeepSeek-V4.1: pack multi-image ViT encoding
- [Perf] kimi-k3.1: Sequence-parallel layer tail for large eager prefill batches
- [Perf] dsa: Scheduled and bf16 MQA logits for the prefill indexer
- [Fix] [PD] Allow bounded backfill past KV-blocked decode requests
- [Bug] Rust extension loader: cargo --version probe can deadlock in launcher spawn context (defensive timeout suggestion)
- [Fix] glm47: Keep string-typed tool-call argument values verbatim
- [Feat] server: Allow cache flushes to retain allocator memory
- Docs
- Python not yet supported