管理大型语言模型(LLM)提示的解决方案

4作者: opuslabs6 个月前原帖
对于那些使用大型语言模型(LLMs)的人来说,你们是如何管理不断增加的提示(如系统提示、模板、链条和实验)在各个项目中的分散情况的?我希望找到一些实用的方法或工具来组织、版本控制和搜索这些提示,以便于它们的精炼和重用,而不是将它们分散在笔记或GitHub片段中。
查看原文
For those of you working with LLMs, how do you manage the growing sprawl of prompts such as system prompts, templates, chains, and experiments across projects? I am looking for practical ways or tools to organize, version, and search prompts so they are easy to refine and reuse, instead of keeping them scattered in notes or GitHub gists.