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
- [Bugfix] Soft-import sgl_kernel_npu in layernorm.py
- [HiSparse]: use logical capacity for request limits
- [disagg] per-rank mooncake transfer engine for multi-HCA nodes
- fix(speculative): default speculative_eagle_topk to 1 for NEXTN/STANDALONE to prevent NoneType crash
- [Bug][diffusion] JoyAI-Image-Edit fails at ImageEncodingStage: Qwen3-VL text encoder missing `mm_token_type_ids`
- [Bug][diffusion] LongLive2 `longlive2_i2v` aborts: num_frames not divisible by num_frames_per_block (causal DMD)
- [model_loader] Stream weights over HTTP Range with one NIC per TP rankPr/object store stream
- [Kernel] DSv4 Q norm-rope (fp8): block-per-token freq share + vectorized dequant for prefill throughput
- [Scheduler] Add a one-sided symmetric-memory path for the DP metadata gather
- :construction: [WIP][llm][npu] Split mixed chunked-prefill FIA into prefill/decode calls on Ascend 950
- Docs
- Python not yet supported