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
- [Perf] [XPU] gemma4: Use native GEMM for narrow BF16 prefill
- [Perf] SenseNova-U1: BSHD q/k/v, fused MLP gate_up + SiLU, RoPE table…
- [Fix] sampling: Refresh is_all_greedy / is_any_greedy when filter_batch removes rows
- [Fix] lora: Clear the embed_tokens/lm_head slot before loading a partial embedding adapter
- [Perf] [Intel] gemma4: Fuse BF16 PLE residual add and layer scale
- [Fix] gemma: Reject input_embeds requests for Gemma 3n and Gemma 4
- [Perf] [Intel] gemma4: Extend BF16 prefill GEMM to intermediate rows
- [Perf] function_call: Stop re-scanning the open DSML invoke on every streamed delta (DeepSeek V3.2/V4/V4.1)
- DeepSeek V3.2/V4/V4.1 streaming tool-call parser: pre-existing bugs found while fuzzing
- [Fix] openai: Forward replayed reasoning to gpt-oss thinking field (#44931, #44932)
- Docs
- Python not yet supported