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
- [Bug] Potential hybrid GDN Radix-cache selected-logprob drift on repeated branch scoring
- [MiniMax-M3] Fix UnboundLocalError in the sparse backend on the MSA path
- [Spec] Fix TypeError on idle batches in the Frozen-KV MTP draft worker
- [NPU] Fix TypeError in the DFlash graph replay: pass global_num_tokens_cpu
- [Bug]: scheduler Req.decoded_text is never written: dead stop-string fallback at schedule_batch.py:1846 and empty DecodeStatus seed on eviction re-init
- [Feature] Save cache from requests and load
- [Feature] Serving VLM VILA
- fix: torchao quantization fails on VLMs
- [Feature] Support `abort_request` in the router
- fix: Prevent errors from being hidden during graph capture with custom all reduce
- Docs
- Python not yet supported