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
- [HiCache] Test device-anchored storage prefetch
- [Bug] MiMo-V2.5: any request with audio hangs the whole server under --enable-dp-attention (audio encoder reduces over the wrong TP group)
- [AMD] DeepSeek-V4: ROCm (gfx950) FP4 sparse indexer
- [Speculative] Reuse Prefill radix prefixes with compact Qwen DFlash
- [Refactor] Simplify RadixCache.total_size() to O(1) using existing size counters
- Fix marlin device-capability check for XPU (None capability)
- [HiSparse] Avoid per-step backup index H2D copy for full batches
- [AMD] Flag when the miles ROCm nightly falls back to a stale image
- Fixed memory leak in decode hicache caused by retraction.
- Fixed memory leak in decode hicache caused by request abort.
- Docs
- Python not yet supported