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] [AMD] dsa: Precompile BF16 sparse MLA splits during warmup
- [Fix] granitemoe: Declare packed_modules_mapping so per-projection quant targets match qkv_proj
- [CI] health-check: Retry the GitHub API before failing a shard
- [Perf] [AMD] GLM5: Add fused KDA prefill backend on gfx950
- [Perf] [XPU] linear: Use torch scaled_mm for MXFP8 linear
- --enable-lmcache with a disabled radix cache crashes the scheduler at first idle (0.5.22)
- [Fix] diffusion/nunchaku: Load FLUX.1 SVDQuant checkpoints again
- [Fix] deepseek-v4.1: size the KV pool by the ratio-1/2 layout's own entry bytes
- [Bug] Glm47MoeDetector.detect_and_parse drops tool calls when the function name is followed by a newline (non-streaming)
- [Feat] dsv4.1: Support CP>1 encoder replay with P/D and HiCache
- Docs
- Python not yet supported