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
- [ROCm] topk v2: split one long row across blocks, the CDNA cluster-path equivalent
- [KDA] Fused intra-chunk prefill path (`chunk_kda_fwd_intra(fuse_diagonal=True)`) collapses for strong per-channel decays because of the ±126 clamp in the exp2 factorization
- fix(fc): enforce parameters schema for named tool_choice in legacy structural tag
- [XPU] Optimize XPU FP8 scales layouts
- [NPU] Add opt-in MLA prefix FIA and Gemma norm fallback
- [Bug] EAGLE spec-decode: num_token_non_padded is 0 for draft batches, masking all draft topk ids to -1 and corrupting the MoE dispatch
- [DeepSeek V4.1] an FP8 Tensor Core path for the FP4 grouped indexer
- Fix EAGLE draft EP dispatch when input_ids is None
- Reuse shared DFlash and DSpark verify metadata planning
- [HiCache] Preserve all MTP KV and recurrent state
- Docs
- Python not yet supported