Prompt Engineering 提示词工程的实践技巧（四）

字数统计: 1.9k字   |   阅读时长≈ 8分

本篇博客主要总结了 DeepLearning AI 公开课程 Prompt Engineering 提示词工程 第五章节 Inferring 推理的内容，也可以理解为该章节的笔记。

输入一些文本，让模型帮助我们完成推理的工作，如提取标签、名字，梳理文本情感等。看起来和之前的输入文本，让模型提取其中的关键信息撰写文章有些类似，不过实际应用中，prompt 还是会有一些不同。

1. 文本情绪分类

传统的机器学习工作流中，要实现文本情绪分类，必须收集标签数据集、训练模型并部署，推断等。对于个人开发者而言，这样的过程工作量大，很繁杂，并不是很合适。LLM将我们从这些繁琐的工作中解放出来，可以通过几个简短的 prompt 就完成这些工作。

在上一个章节中，我们会采用如下的 prompt 来指示 LLM，让它提取文本。

1
2
3
4
5
6
7
8
9
10
lamp_review = """
Needed a nice lamp for my bedroom, and this one had additional storage and not too high of a price point. Got it fast. The string to our lamp broke during the transit and the company happily sent over a new one. Came within a few days as well. It was easy to put together. I had a missing part, so I contacted their support and they very quickly got me the missing piece! Lumina seems to me to be a great company that cares about their customers and products!!
"""
prompt = f"""
What is the sentiment of the following product review, which is delimited with triple backticks?

Review text: '''{lamp_review}'''
"""
## 输出
The sentiment of the review is positive. The reviewer is satisfied with the lamp they purchased, mentioning the additional storage, reasonable price, fast delivery, good customer service, and ease of assembly. They also praise the company for caring about their customers and products.

可以看到，虽然正确提取了信息，但是不够凝练。对prompt做一些改变，要求输出更凝练：

1
2
3
4
5
6
7
8
9
10
prompt = f"""
What is the sentiment of the following product review, 
which is delimited with triple backticks?

Give your answer as a single word, either "positive" or "negative".

Review text: '''{lamp_review}'''
"""
## 输出
Positive

可以看到，要求LLM对情感进行二元分类，并给出固定的选项，可以让回复更标准，更易于在程序之间传递。

2. 文本情绪识别

更进一步，为了更细致地了解文本中包含的情绪，我们可以让LLM将文本中的多种情绪概括成一个词语列表，如下所示：

1
2
3
4
5
6
7
prompt = f"""
Identify a list of emotions that the writer of the following review is expressing. Include no more than five items in the list. Format your answer as a list of lower-case words separated by commas.

Review text: '''{lamp_review}'''
"""
## 输出
happy, satisfied, grateful, impressed, content

和前面我们要求 LLM 判断这段文本的态度类似，我们可以要求 LLM 帮助我们识别出特别不满的情绪，并返回 YES 或 NO。

1
2
3
4
5
6
7
prompt = f"""
Is the writer of the following review expressing anger? The review is delimited with triple backticks. Give your answer as either yes or no.

Review text: '''{lamp_review}'''
"""
## 输出
No

我们还可以让 LLM 帮助我们提取评论中涉及的产品和制造商，并以JSON的格式返回。这在批量处理用户评价、反馈时常常用到。例如在游戏中，收集某个活动期间玩家的公平发言，并分析是否有对某个活动极其不满的情绪，我们就可以让 LLM 提取表达不满的玩家所提到的活动名字。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
prompt = f"""
Identify the following items from the review text: 
- Item purchased by reviewer
- Company that made the item

The review is delimited with triple backticks. Format your response as a JSON object with 
"Item" and "Brand" as the keys. If the information isn't present, use "unknown" as the value. Make your response as short as possible.
  
Review text: '''{lamp_review}'''
"""
## 输出

{
  "Item": "lamp",
  "Brand": "Lumina"
}

3. 一次性处理多个任务

将prompt整合，我们可以将上述的多个任务，通过一次请求完成。在指示中，一定要足够清晰、明确，比如指出返回的选项、数据类型，要求返回JSON，并指明 key 值。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
prompt = f"""
Identify the following items from the review text: 
- Sentiment (positive or negative)
- Is the reviewer expressing anger? (true or false)
- Item purchased by reviewer
- Company that made the item

The review is delimited with triple backticks. Format your response as a JSON object with "Sentiment", "Anger", "Item" and "Brand" as the keys.If the information isn't present, use "unknown" as the value. Make your response as short as possible. Format the Anger value as a boolean.

Review text: '''{lamp_review}'''
"""
## 输出
{
    "Sentiment": "positive",
    "Anger": false,
    "Item": "lamp", # 视频中展示的输出比我自己运行得到的要更详细一些
    "Brand": "Lumina"
}

4. 文本主题推断

LLM 的另一个技能是推断主题。

提炼文本主题

输入一段长文本，我们可以要求 LLM 确定这段文本所讨论的主题。我们可以指定 LLM 提炼几个主题，每个主题用一到两个单词描述，以逗号分隔的列表来表示。由于示例文本比较长，在此不另外列出，实际上，用什么样的文本作为输入并不是很重要。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
prompt = f"""
Determine five topics that are being discussed in the following text, which is delimited by triple backticks.

Make each item one or two words long. 

Format your response as a list of items separated by commas.

Text sample: '''{story}'''
"""
## 输出
government survey, job satisfaction, NASA, Social Security Administration, employee concerns

## 使用split输出
response.split(sep=',')
['government survey', ' job satisfaction', ' NASA', ' Social Security Administration', ' employee concerns']

文本主题索引

既然 LLM 能够提取出一段文本中的主题，那么反过来，我们也可以让 LLM 判断，某个主题是否为文本的主题之一。

首先，我们要整理出我们关心的主题 list。将这个主题 list 和文本都输入给 LLM，要求 LLM 根据主题是否相关，以 0/1 的形式表示结果。可以注意到，在prompt中，我们将 topic_list 的格式转化为了一个逗号分隔的文本形式。因为我们不能直接将python定义的数组嵌入 prompt 中。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
topic_list = ["nasa", "local government", "engineering", "employee satisfaction", "federal government"]

prompt = f"""
Determine whether each item in the following list of topics is a topic in the text below, which is delimited with triple backticks.

Give your answer as follows: item from the list: 0 or 1

List of topics: {", ".join(topic_list)}

Text sample: '''{story}'''
"""
## 输出
nasa: 1
local government: 0
engineering: 0
employee satisfaction: 1
federal government: 1

这个文本主题推断的技能，还可以稍加包装，用作文章提醒。例如，定义一个你感兴趣的主题 list，就像前文的 topic_list一样，每次有新增的文章发布时，就执行一遍上面的操作。在拿到返回值以后，根据返回的 0/1，判断是否要给用户发送提醒。这其实就可以成为文章订阅机制的一部分。

1
2
3
topic_dict = {i.split(': ')[0]: int(i.split(': ')[1]) for i in response.split(sep='\n')}
if topic_dict['nasa'] == 1:
    print("ALERT: New NASA story!")

Prompt Engineering 提示词工程笔记汇总

参考资料

ChatGPT Prompt Engineering for Developers - DeepLearning.AI

B站可以找到搬运翻译的中文版本👆：《吴恩达 x OpenAI Prompt课程》

赏

谢谢你请我吃糖果

支付宝
微信

扫一扫，分享到微信

tag:

    缺失模块。
    1、请确保node版本大于6.2
    2、在博客根目录（注意不是yilia-plus根目录）执行以下命令：
    npm i hexo-generator-json-content --save

    3、在根目录_config.yml里添加配置： 

      jsonContent:
    meta: false
    pages: false
    posts:
      title: true
      date: true
      path: true
      text: false
      raw: false
      content: false
      slug: false
      updated: false
      comments: false
      link: false
      permalink: false
      excerpt: false
      categories: false
      tags: true

一个游戏行业的开发小白。什么都学了一些，什么都会一点。AIGC/ComfyUI工作流搭建/AI聊天机器人/Lua/C++……欢迎咨询交流。欢迎大家指点交流~ 邮箱: wuxinbei0409@gmail.com