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
- [DSv4.1][SM90] Fuse inverse RoPE into a single-launch WO-A kernel for decode
- [Bug] Ordinary API key can clear persistent HiCache storage
- [Fix] deterministic: Keep the GDN output tile fixed for varlen prefill
- [Fix] GDN: Take prefix-cache checkpoints from the fp32 state in the Triton kernel
- [Perf] Qwen-Image 2.1: fused gate/up + SwiGLU MLP on comfy_kitchen INT8
- [Fix] [AMD] mamba: Import the causal conv1d kernels for Mamba2 layers on ROCm
- [Fix] [AMD] activation: Run ReLU2 natively on ROCm
- Scheduler: broadcast TP requests via the shared-memory Message Queue instead of gloo
- [Fix] multimodal: Keep per-request images when a batch mixes image and text-only requests
- [Fix] diffusion: Invalidate merged LoRA cache after adapter rewrites
- Docs
- Python not yet supported