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
- [Chore] deps: Bump FlashInfer to 0.7.4
- [Fix] scheduler: Answer release_memory_occupation only after every pipeline stage released
- [Feat] offloader: Let OffloaderV1 offload only the model-declared submodule
- [Fix] rust-server: Log the drain's scheduler kill at info and bound the reap
- [Perf] [NPU] sampling: skip full-vocabulary un-sort in top-k renorm
- [Fix] [AMD] docker: Pin UCX and NIXL commits for ROCm PD disaggregation
- [Feat] eplb: Add ring-buddy expert replication (--ep-dispatch-algorithm ring) for Muse Spark
- [Feat] [NVIDIA] multimodal: Honor SGLANG_FORCE_CPU_IMAGE_PREPROCESSING in the base multimodal processor
- [Fix] [NVIDIA] moe: Stop DeepGEMM's MLOPart helper once MegaMoE creates its buffer
- [Bug] moe_route_warp crashes on CPU-offloaded correction_bias with --cpu-offload-gb (Tensor match failed ... device=cpu at route_warp.cuh:459)
- Docs
- Python not yet supported