➡️
继续阅读
-
微调 vs. 冻结嵌入:何时该调整 LLM,何时该调整 Sklearn 分类头
文章比较两种领域适配LLM策略:冻结嵌入加scikit-learn分类器,与QLoRA微调。在AG News四分类实验中,前者提取嵌入约96.5秒、分类器...
-
BabyLog记录了我14年对孩子所有的回忆也见证了我自学编程的学习路线
作者回顾14年自学编程历程,为记录孩子成长开发BabyLog:从照书敲PHP起步,历经ThinkPHP、SpringBoot、Flask、前后端分离,再到...
-
从参数对称性到有效学习率
近来,越来越多的论文或者博客,开始基于参数的对称性来分析和指导模型的优化,比如Hyperball、HyperTransfer、ELR等,它们总体都是围绕着...
-
How to Extract Data from InnoDB, MyISAM, and MyRocks Files Without a Running Server
Probably everyone responsible for managing database servers at some point rea...
-
Streamline creation and management of AI-ready, custom apps with SharePoint Embedded
SharePoint content isn’t limited to sites and document libraries. Increasingl...
-
Fragments: October 28
Andrej Karpathy sees a widening gap in how people see LLM capabilities. The ...