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
- [Feat] dsv4: Serve routed experts as MXFP8 on a block-wise FP8 checkpoint
- [Fix] weight-checker: Skip non-persistent axis_map buffer in reset_tensors
- [Refactor] trtllm: derive TRTLLMHAAttnBackend from AttentionBackend directly
- [Perf] [NVIDIA] moe: Skip padded DSpark EP8 CUTLASS rows on Hopper
- [Perf] [NVIDIA] deepseek-v4: Enable Hopper G8 token-major WO-A verify
- [Perf] [NVIDIA] dp-attention: Use short Hopper DPA8 collective paths
- nightly-test-amd.yml reports "All ROCm nightly test jobs passed" when 46 of 48 jobs were skipped
- [Fix] benchmark: Keep the image placeholder token out of random image prompts for processor-less checkpoints
- [Chore] deps: Bump xgrammar to 0.2.9
- [Fix] lora: Report adapter weights that should_apply_lora leaves unapplied
- Docs
- Python not yet supported