首页 > 其他分享 >斯坦福小镇prompt_templete学习小结

斯坦福小镇prompt_templete学习小结

时间:2023-12-11 10:34:11浏览次数:40  
标签:templete persona prompt name -- conversation init 小结 target

提取要点:

简单说就是,每个行为都有一定模板,给gpt生成 

 

[Conversation]

All convo utterances  全部对话话语

Write down if there is anything from the conversation that !<INPUT 1>! might have found interesting from !<INPUT 2>!'s perspective, in a full sentence.

 

Here is the memory that is in !<INPUT 1>!'s head:

!<INPUT 2>!

Past Context:

!<INPUT 3>!

Task: Given the above, what should !<INPUT 9>! say to !<INPUT 10>! next in the conversation? And did it end the conversation?

 

Here is the conversation that happened between !<INPUT 0>! and !<INPUT 1>!.

!<INPUT 2>!

Summarize what !<INPUT 3>! thought about !<INPUT 4>! in one short sentence. The sentence needs to be in third person:

 

 

!<INPUT 12>! and !<INPUT 13>! are in !<INPUT 14>!. What would they talk about now?

 

 

原文:

对话变成记忆

memo_on_convo_v1.txt

Variables:

!<INPUT 0>! -- All convo utterances

!<INPUT 1>! -- persona name

!<INPUT 2>! -- persona name

!<INPUT 3>! -- persona name

 

<commentblockmarker>###</commentblockmarker>

[Conversation]

!<INPUT 0>!

 

Write down if there is anything from the conversation that !<INPUT 1>! might have found interesting from !<INPUT 2>!'s perspective, in a full sentence.

 

"!<INPUT 3>!

 

 

发起对话

iterative_convo_v1.txt

 

Variables:

!<INPUT 0>! -- persona ISS

!<INPUT 1>! -- persona name

!<INPUT 2>! -- retrieved memory

!<INPUT 3>! -- past context

!<INPUT 4>! -- current location

!<INPUT 5>! -- current context

!<INPUT 6>! -- persona name

!<INPUT 7>! -- target persona name

!<INPUT 8>! -- curr convo

!<INPUT 9>! -- persona name

!<INPUT 10>! -- target persona name

!<INPUT 11>! -- persona name

!<INPUT 12>! -- persona name

!<INPUT 13>! -- persona name

<commentblockmarker>###</commentblockmarker>

Context for the task:

 

PART 1.

!<INPUT 0>!

 

Here is the memory that is in !<INPUT 1>!'s head:

!<INPUT 2>!

 

PART 2.

Past Context:

!<INPUT 3>!

 

Current Location: !<INPUT 4>!

 

Current Context:

!<INPUT 5>!

 

!<INPUT 6>! and !<INPUT 7>! are chatting. Here is their conversation so far:

!<INPUT 8>!

 

---

Task: Given the above, what should !<INPUT 9>! say to !<INPUT 10>! next in the conversation? And did it end the conversation?

 

Output format: Output a json of the following format:

{

"!<INPUT 11>!": "<!<INPUT 12>!'s utterance>",

"Did the conversation end with !<INPUT 13>!'s utterance?": "<json Boolean>"

}

 

对话变成想法

convo_to_thoughts_v1.txt

 

Variables:

!<INPUT 0>! -- init persona name

!<INPUT 1>! -- target persona name

!<INPUT 2>! -- convo string

!<INPUT 3>! -- init persona name

!<INPUT 4>! -- target persona name or "the conversation"

 

<commentblockmarker>###</commentblockmarker>

Here is the conversation that happened between !<INPUT 0>! and !<INPUT 1>!.

 

!<INPUT 2>!

 

Summarize what !<INPUT 3>! thought about !<INPUT 4>! in one short sentence. The sentence needs to be in third person:

 

 

对话中

create_conversation_v2.txt

 

Variables:

!<INPUT 0>! -- init_persona iss

!<INPUT 1>! -- target_persona iss

 

!<INPUT 2>! -- init_persona_name

!<INPUT 3>! -- target_persona_name

!<INPUT 4>! -- init_persona's thoughts

 

!<INPUT 5>! -- target_persona_name

!<INPUT 6>! -- init_persona_name

!<INPUT 7>! -- target_persona's thoughts

 

!<INPUT 8>! -- current time

!<INPUT 9>! -- init_persona curr action description

!<INPUT 10>! -- target_persona curr action description

 

!<INPUT 11>! -- previous convo

 

!<INPUT 12>! -- init_persona_name

!<INPUT 13>! -- target_persona_name

!<INPUT 14>! -- curr_location name

!<INPUT 15>! -- init_persona_name

 

<commentblockmarker>###</commentblockmarker>

We have two characters.

 

Character 1.

!<INPUT 0>!

 

Character 2.

!<INPUT 1>!

---

Context:

Here is what !<INPUT 2>! thinks about !<INPUT 3>!:

!<INPUT 4>!

Here is what !<INPUT 5>! thinks about !<INPUT 6>!:

!<INPUT 7>!

Currently, it is !<INPUT 8>!

-- !<INPUT 9>!

-- !<INPUT 10>!

!<INPUT 11>!

 

!<INPUT 12>! and !<INPUT 13>! are in !<INPUT 14>!. What would they talk about now?

 

!<INPUT 15>!: "

标签:templete,persona,prompt,name,--,conversation,init,小结,target
From: https://www.cnblogs.com/cosmowind/p/17893811.html

相关文章

  • windows下的flex小结
    总结flex是一个词法分析程序自动生成工具。自动生成工具flex生成给定的DFA所对应的语言的词法分析程序,对每个状态编写对应的识别代码,然后对每个状态的代码用合适的分支、循环等语句进行整合,完成整个代码的合成工作,最后会生成一个CSourceFile程序,即为所要求的词法分析程序。在......
  • 词法分析小结
    总结编译程序时,第一步要做的就是词法分析。用词法分析对输入的字符串形式的源程序按顺序进行扫描,针对指定的语言,根据源程序的词法规则识别具有独立意义的单词(符号),其中对于错误的词法也要进行识别,并输出与其等价的Token序列。首先,构造识别单词的自动机,例如:其次,将各类单词的状......
  • 【EMNLP 2023】面向Stable Diffusion的自动Prompt工程算法BeautifulPrompt
    近日,阿里云人工智能平台PAI与华南理工大学朱金辉教授团队合作在自然语言处理顶级会议EMNLP2023上发表了BeautifulPrompt的深度生成模型,可以从简单的图片描述中生成高质量的提示词,从而使文生图模型能够生成更美观的图像。BeautifulPrompt通过对低质量和高质量的提示进行微调,并进一步......
  • Prompt创新引领对话系统未来
    随着人工智能技术的快速发展,自然语言处理领域中的对话系统成为了研究热点。千帆大模型平台的海量Prompt模板库,给该领域带来了前所未有的变革和影响。本文将重点探讨千帆大模型平台的海量Prompt模板库的三个主要特点及其实践意义。一、海量数据,丰富多样的对话系统训练素材千帆大模型......
  • 掌握Prompt工程,开启AI大模型新时代
    人工智能(AI)领域正在快速发展,各种专业术语和黑话也层出不穷,使得新入行的人往往感到无所适从。特别是在涉及到AI大模型的领域,一些关键概念如提示工程(prompt)、向量工程(embedding)和微调工程(fine-tune)更是晦涩难懂。本文将用简洁明了的语言,详解这些行业黑话,帮助你迅速搞懂它们。一、提示......
  • git使用小结
    一、提交代码到master仓库1、克隆新建仓库地址gitclone"xxx"2、查看仓库状态gitstatus3、拷贝上传的代码到目录并上传gitadd.4、加备注,加说明gitcommit-m"firstcommit"5、输入用户名和邮箱gitconfig--globaluser.name"wanggangtao"gitconfig--globalus......
  • 解密Prompt系列20. LLM Agent之再谈RAG的召回多样性优化
    几个月前我们就聊过RAG的经典方案解密Prompt系列14.LLMAgent之搜索应用设计。前几天刚看完openAI在DevDay闭门会议上介绍的RAG相关的经验,有些新的感悟,借此机会再梳理下RAG相关的优化方案。推荐直接看原视频(外网)ASurveyofTechniquesforMaximizingLLMPerformanceRAG最关键......
  • 第 132 场周赛——质数小结论,并查集配Floyd
    https://www.acwing.com/activity/content/competition/problem_list/3648/B题收获:1.利用题目告诉的结论:1e9范围质数之差小于3002.一个数不被2-a的任何数整除等价于他的最小质因子需要大于ac题:初步宏观思路:不难想到用并查集维护类别,只需将每一个类缩成一个点,由于最多只有500......
  • Python正则表达式(小结)
    1.什么是正则表达式?正则表达式(regularexpression,有时简写为RegEx或regex)就是用一组由字母和符号组成的“表达式”来描述一个特征,然后去验证另一个“字符串”是否符合/匹配这个特征。2.应用场景?(1)验证字符串是否符合指定特征,比如验证邮件地址是否符合特定要求等;(2)用来查找字......
  • 各类运算符的小结
    packageoperator;publicclassdemo01{publicstaticvoidmain(String[]args){inta=111;intb=1213;intc=1131;System.out.println((a+b+c)*a);System.out.println(a/(double)b);doubled=11111;doublee......