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] hicache: Log and label buffer-only backup drops
- [Perf] Gather the striped prefix KV for DSA extend under DCP (env-gated prototype)
- [Feat] QSA: Integrate compressed QSA with the unified hybrid memory pool
- [Feat] [AMD] kimi_k3: Add whole-layer fused KDA backend (gfx950 TP8)
- [Feat] beam-search: Add trie head and shared-context attention
- [Refactor] [NVIDIA] attention: Migrate DSA/DSV4 indexers to DeepGEMM 0.2.5.post2
- [Feat] lora: Add merge and unmerge endpoints
- [Perf] DSA: Share one metadata buffer set across MTP draft steps
- [Feat] [NVIDIA] flashinfer_megamoe: SM120 prefill support with FlashInfer's cake MegaMoE backend (DeepSeek-V4-Flash)
- [Perf] [AMD] kimi-k3: Run the merged MoE front through AITER tuned BF16 GEMM
- Docs
- Python not yet supported