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
- [Diffusion] fp8_fa_sm120: keep listed attention layers in BF16
- [Bug] Gemma4ForCausalLM loader ignores global_head_dim for full-attention layers — text-only Gemma‑4 A4B checkpoints fail to load (same weights load fine as Gemma4ForConditionalGeneration)
- [Bug] PPLX MoE A2A padded tokens with topk_ids = -1
- [PD] Add UniFlow as a native KV transfer backend
- [Fix] attn_residual: int64 program ids in the score/combine kernels (int32 overflow at 37,451 tokens)
- [Platform] Unify multimodal_gen platforms onto DeviceMixin
- [Fix] GGUF: make the MMQ branch of fused_moe_gguf runnable
- [Bugfix] DeepSeek V4: fix mHC prewarm when chunked prefill is disabled
- [Feature] Capture aux hidden states on the target without a speculative draft
- [Feature] Forward observer and deferred outputs for out-of-tree consumers
- Docs
- Python not yet supported