Enable javascript in your browser for better experience. Need to know to enable it? Go here.
Published : Sep 27, 2023
NOT ON THE CURRENT EDITION
This blip is not on the current edition of the Radar. If it was on one of the last few editions, it is likely that it is still relevant. If the blip is older, it might no longer be relevant and our assessment might be different today. Unfortunately, we simply don't have the bandwidth to continuously review blips from previous editions of the Radar. Understand more
Sep 2023
Assess ?

GPTCache 是一个用于大型语言模型(LLM)的语义缓存库。我们认为需要在 LLM 前增设缓存层主要出于两种原因——通过减少外部 API 调用来提升整体性能,以及通过缓存近似响应来减少运营成本。不同于使用精确匹配的传统缓存方式, 基于 LLM 的缓存解决方案需要对输入进行相似或相关匹配。 GPTCache 通过使用嵌入算法将输入转化为嵌入,再通过向量数据库对这些嵌入进行相似性搜索。这种设计有一个缺点,可能会导致缓存命中时遇到假阳性结果,或缓存未命中时遇到假阴性结果,因此我们建议你在构建基于 LLM 应用时,仔细评估 GPTCache。

Download the PDF

 

 

 

English | Português 

Sign up for the Technology Radar newsletter

 

 

Subscribe now

Visit our archive to read previous volumes