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
- [Fix] utils: Fix info_once message deduplication, is_same_type symmetry, and trim_overlap slicing
- [Fix] hicache: Keep write-back admission from promising device KV that eviction cannot free
- [Perf] dsa: Enable K-only short prefill with PCP
- [Test] mem_cache: Skip DSA indexer sharding test below SM 8.9
- [Feat] [Intel] monokernel: --enable-xpu-monokernel, the Intel XPU backend of the monokernel binding
- [Perf] pp: Send decode activations before CPU result processing
- [Fix] mamba: Preserve completed checkpoints across overlap gaps
- [Fix] benchmark: reject a range ratio outside [0, 1]
- [Fix] deepseek-v4.1: Keep active DP decode ranks out of idle tail replay
- Post-hoc Req.finished_len mutation desyncs pool accounting — invariant checker fires with negative residue (undocumented contract)
- Docs
- Python not yet supported