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
- [Perf] [AMD] mimo-v2.6: Fuse RoPE and KV-cache insertion on gfx950
- [Fix] moe: Sync precompile lock and configure deepep_v2 timeouts
- [Fix] benchmark: Count actual overlaps for peak concurrency
- [Fix] speculative: FR-Spec (--speculative-token-map) with a quantized (FP8) target lm_head
- [Bug] benchmark peak concurrency counts sequential requests as overlapping
- [Feat] [AMD] mori: Add Fabric PD backend for rack-level gfx1250 serving
- [Feat] simulator: Support closed-loop agentic workloads
- [Bug] EP2 (tp=2 + ep=2) crashes with illegal memory access on Qwen3.8-Flash-Next-NVFP4 (RTX PRO 6000, sm_120)
- [Feat] cp: Extend GLM prefill BCG support to EP and NIXL PD disaggregation
- [Fix] auth: Require admin API key for flush_cache and LoRA endpoints
- Docs
- Python not yet supported