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
- [CP] Support interleave prefill for dense MHA, GQA and MLA
- [CP] Support zigzag prefill for DeepSeek sparse attention
- [CP] Support CUDA zigzag prefill for DeepSeek V4
- [Diffusion] Add Wan2.2-S2V-14B speech-to-video (port of #21609)
- [Fix] Honor Anthropic thinking settings for MiniMax-M3
- fix(args): floor the derived mem_fraction_static and validate explicit values
- [Diffusion] Wan2.2-S2V: wav2vec2 lip-sync fix and DiT speedups (follow-up to #43279)
- [Spec] Declare kv_indptr/kv_indices on EagleVerifyInput for DP-attention idle verify
- [Fix] Preserve text sub-config dtype under thinker_config in get_hf_text_config
- [MegaMoE] W4A4 (mxf4xmxf4) target path for Kimi-K3: feature-completion patches + bs=1 perf/numerics findings
- Docs
- Python not yet supported