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
- [HiCache] Issue round-head H2D copies before the load burst hand-over
- [Bug] UnifiedRadixCache: L3 storage is never consulted when a prefix survives only as backuped stubs after host-tier eviction
- [Perf] Persistent lazy eviction heaps for the unified radix cache full component
- [Core] Load plugins before ServerArgs resolves the model, and in every spawned process
- [Feature] HiCache: run the host->device load-back enqueue on a loader thread so it can overlap the forward
- [Feature] Support pre-sampled video input across native and encoder paths
- [HiCache] Run the host->device load-back enqueue on a loader thread (opt-in)
- [kv-shard 4/4] PD transfer and e2e test
- test: add GLM-5.3-Flash HiCache KL coverage
- [DSV4] Use mhc_pre's split-K pre-norm GEMM in the mhc_fused_post_pre fallback
- Docs
- Python not yet supported