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] sgl-router: set TCP_NODELAY on accepted client connections
- [Perf] beam: Add the SM103 single-token BF16 MoE GEMV follow-up
- [Perf] lora: Eliminate redundant dict key copies and slot rescans in LoRA eviction and batch prep
- [Feat] [CPU] cpu-graph: Support TorchInductor cpp_wrapper with direct custom op C shims
- [Fix] environ: restore the env var when an override block raises
- [Refactor] multimodal: [MM refactor 2/9] Initialize multimodal processors from a processor config
- [Refactor] multimodal: [MM refactor 3/9] Extract a shared multimodal processor mixin
- [Bug] JIT link fails with the pip CUDA toolkit: -L$CUDA_HOME/lib64 -lcudart doesn't match its lib/libcudart.so.13 layout
- [Fix] jit: Link libcudart from the pip CUDA toolkit layout
- [Fix] qwen4_exp: Split the PLE prefill short conv under the int32 limit
- Docs
- Python not yet supported