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
- fix(causal_conv1d_triton): cast conv_states loads to compute dtype to fix Triton type mismatch
- feat(multimodal_gen): add --max-queued-requests backpressure
- [Help] [Performance] PD disaggregation on H200 shows no throughput gain over single-node traditional deployment with 32k input / 512 output
- [fix] upstreams for qwen3_5.py and qwen3.5 lora on qkv layers fix.
- [NPU] Support Longcat Flash for NPU
- Cherry pick sglang 20114 22940 23427
- [LoRA][MoE] NVFP4 LoRA on FlashInfer-CUTLASS MoE backend
- [RFC][Refactor] `mem_cache` pool / allocator restructure
- [1/4][LoRA] Add NVFP4 Cutlass MoE LoRA runner
- [2/4][LoRA] Avoid NVFP4 Cutlass LoRA layout bridges
- Docs
- Python not yet supported