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
- Qwen3.8-Flash-Next Roadmap
- Fix draft KV memory sizing with DP attention
- fix: make scripted runtime request ID reuse and abort reliable
- [Bug] Qwen3.5 hybrid (GDN) GPTQ checkpoint: `linear_attn.in_proj_ba` built as quantized although the checkpoint stores `in_proj_a/b` as bf16 → 96× "not found in params_dict", then `gptq_marlin_repack` fails (size_n=96)
- [HiCache] Honor host duplicate reclaim opt-out in Rust TreeCore
- [Rust] Extract shared environment parsing helpers
- [Bug] DeepSeek V4/V3.2 DSML tool-call parser occasionally wraps arguments in a spurious "arguments"/"input" key
- Fix NIXL backend device selection in initialization thread
- [Bug] sgl_kernel flash_attn: is_fa3_supported() accepts sm_89 but no sm_89 cubin ships, and `ver` is ignored
- Refactor shared cache transfer types
- Docs
- Python not yet supported