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
- [CPU][Spec] Prepack the DFlash draft context projection for AMX
- [SM120] Pad small-M in_proj_ba GEMMs in Qwen3.5 GDN past a cuBLAS heuristic cliff
- [MoE][miles] Cherry-pick #33743: fix flashinfer TRT-LLM BF16 expert weight reload on refit
- [kernel] Add tuned fused_moe config for NVIDIA RTX PRO 5000 Blackwell (SM120), E=288/N=256 (corrected key)
- [Fix] Normalize Gemma3n attention output before o_proj
- [CI] Test Escape Report: /rerun-test did not enforce registered test __main__ validation
- fix(humming): support CP decode attention TP for GLM W4AFP8
- Reduce KDA prefill OOM risk with opt-in workspace limits
- Free raw image tensors before language-model execution to avoid prefill OOMs
- Allow CPU image preprocessing to preserve serving GPU memory
- Docs
- Python not yet supported