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
- [Bug] Rust server assigns the same HTTP port to multiple system-DP replicas
- test(npu): add e2e smoke test for --radix-eviction-policy-config
- [Bug] FP8 KV cache scales are silently dropped when loading Qwen3-MoE / Qwen2 / Mixtral and every layer falls back to scale 1.0
- Fix fp8 KV cache scale loading for Qwen3-MoE / Qwen2 / Mixtral
- [mem_cache] Add a per-cache-salt TTL for radix prefixes, with optional KV zeroization
- fix(retract): restart output cursors when input_embeds generation restarts
- [model-gateway] Release prefill worker load when the prefill response arrives
- [diffusion] attention: Wire fast path varlen attention through the attention backends
- [RFC] Asymmetric P/D deployment for DeepSeek-V4.1 Flash
- [Model] Support BerryLM (hybrid KDA MoE with Gated Block AttnRes)
- Docs
- Python not yet supported