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
- Fix duplicate output on incremental streaming aborts
- [Bug] DeepSeek-V4 + HiCache write_through: TP ranks deadlock under concurrent long prefills (scheduler and detokenizer go silent, /health 503)
- Fix DeepSeek V3.1 parsing of unwrapped structural tool calls
- Fix Cohere Command4 tool calls coalesced with streaming text
- [AMD] Deselect slow survives_real_pytest_runner from diffusion unit step
- [Spec] Keep the speculative lookahead within context_len at the length cap
- fix(parser): preserve Harmony tool arguments across chunks
- [Router] Respect cache_salt in KV-aware routing
- [Perf] Native tensor output and prefill scoring optimizations
- [Test] diffusion: Cover CUDA and HIP latency guard policies
- Docs
- Python not yet supported