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
- [CI] mscclpp: Introduce CI for MSCCL++ all-reduce on the single-node 8-GPU H200 runner
- [Fix] diffusion: Apply RoPE to the FLUX.2-dev negative CFG branch
- [Feat] spec: support adaptive speculative decoding for STANDALONE worker
- [Fix] [XPU] sampling: Make deterministic top-k renorm reproducible on XPU
- [Test] cp: Re-enable the CP8 TP-decode server test with measured logprob bounds
- [CI] [AMD] amd-ci: Pass optional HF proxy env into the CI container
- [Fix] attention: Use int64 token offsets in layernorm_gated and fused_qk_rmsnorm_rope_gate
- [Fix] debug_utils: Stop grafter tests hanging on a reused port
- [Feat] [NVIDIA] weight-sync: Add GPU-delta in-place weight updates
- [Feat] dflash: Shared context K/V drafts with a one-layer draft KV pool
- Docs
- Python not yet supported