20260625:很多新内容
This commit is contained in:
16
concepts/sglang.md
Normal file
16
concepts/sglang.md
Normal file
@@ -0,0 +1,16 @@
|
||||
---
|
||||
title: "SGLang"
|
||||
created: 2026-06-24
|
||||
updated: 2026-06-24
|
||||
type: concept
|
||||
tags: ["inference-engine", "llm-serving"]
|
||||
sources:
|
||||
- "[[unlimited-ocr-works-2026]]"
|
||||
---
|
||||
|
||||
# SGLang
|
||||
|
||||
SGLang 是高效的 LLM 推理引擎。Unlimited OCR 为其实现了 R-SWA 的 KV cache 管理支持和优化,使模型在 SGLang 上能以恒定 TPS 和 GPU 内存运行。
|
||||
|
||||
## 参考
|
||||
- [[unlimited-ocr-works-2026]]
|
||||
Reference in New Issue
Block a user