Prompt Engineering

ChatGPT 的原理

基本原理：LLM produces a probability distribution over some vocabulary

GPT 的训练目标只是预测下一个最有可能出现的单词是什么，但是不知道怎样回答问题。InstructGPT 通过在一系列 QA 对话上进行对齐训练（RLHF），帮助 GPT 理解怎样回答问题。

如何设置 openAI 的 API 接口

import openai
import os

# Get the value of 'OPENAI_API_KEY' environment variable
openai.api_key  = os.getenv('OPENAI_API_KEY')

# Helper function to get the response from ChatGPT
def get_completion(prompt, model="gpt-3.5-turbo"):
    messages = [{"role": "user", "content": prompt}]
    response = openai.ChatCompletion.create(
        model=model,
        messages=messages,
        temperature=0, # degree of randomness of the model's output
    )
    return response.choices[0].message["content"]

Prompting 原则 1：指令要清晰且具体

技巧 1：使用分隔符来清晰划分输入文本的不同部分：

text = f"""
You should express what you want a model to do by \
providing instructions that are as clear and \
specific as you can possibly make them. \
This will guide the model towards the desired output, \
and reduce the chances of receiving irrelevant \
or incorrect responses. Don't confuse writing a \
clear prompt with writing a short prompt. \
In many cases, longer prompts provide more clarity \
and context for the model, which can lead to \
more detailed and relevant outputs.
"""
prompt = f"""
Summarize the text delimited by triple backticks \
into a single sentence.
```{text}```
"""
response = get_completion(prompt)
print(response)

技巧 2：对 ChatGPT 的回答进行结构化

prompt = f"""
Generate a list of three made-up movie titles along \
with their authors and genres.
Provide them in JSON format with the following keys:
book_id, title, author, genre.
"""
response = get_completion(prompt)
print(response)

技巧 3：让模型先判断文本是否满足特定的条件，再做出相应的回答

text_1 = f"""
Making a cup of tea is easy! First, you need to get some \
water boiling. While that's happening, \
grab a cup and put a tea bag in it. Once the water is \
hot enough, just pour it over the tea bag. \
Let it sit for a bit so the tea can steep. After a \
few minutes, take out the tea bag. If you \
like, you can add some sugar or milk to taste. \
And that's it! You've got yourself a delicious \
cup of tea to enjoy.
"""
prompt = f"""
You will be provided with text delimited by triple quotes.
If it contains a sequence of instructions, \
re-write those instructions in the following format:

Step 1 - ...
Step 2 - …
…
Step N - …

If the text does not contain a sequence of instructions, \
then simply write \"No steps provided.\"

\"\"\"{text_1}\"\"\"
"""
response = get_completion(prompt)
print("Completion for Text 1:")
print(response)

技巧 4：”Few-shot”指令：给模型一个成功案例去模仿

prompt = f"""
Your task is to answer in a consistent style.

<child>: Teach me about patience.

<grandparent>: The river that carves the deepest \
valley flows from a modest spring; the \
grandest symphony originates from a single note; \
the most intricate tapestry begins with a solitary thread.

<child>: Teach me about resilience.
"""
response = get_completion(prompt)
print(response)

Prompting 原则 2：给模型更多思考的时间

技巧 1：把复杂的任务分解成一个个具体的小任务，并告诉模型每一步要怎么做。

text = f"""
In a charming village, siblings Jack and Jill set out on \
a quest to fetch water from a hilltop \
well. As they climbed, singing joyfully, misfortune \
struck—Jack tripped on a stone and tumbled \
down the hill, with Jill following suit. \
Though slightly battered, the pair returned home to \
comforting embraces. Despite the mishap, \
their adventurous spirits remained undimmed, and they \
continued exploring with delight.
"""
prompt = f"""
Your task is to perform the following actions:
1 - Summarize the following text delimited by
  <> with 1 sentence.
2 - Translate the summary into French.
3 - List each name in the French summary.
4 - Output a json object that contains the
  following keys: french_summary, num_names.

Use the following format:
Text: <text to summarize>
Summary: <summary>
Translation: <summary translation>
Names: <list of names in Italian summary>
Output JSON: <json with summary and num_names>

Text: <{text}>
"""
response = get_completion(prompt)
print(response)

技巧 2：指导模型在快速得出结论之前先探索自己的解决方案

prompt = f"""
Your task is to determine if the student's solution \
is correct or not.
To solve the problem do the following:
- First, work out your own solution to the problem.
- Then compare your solution to the student's solution \
and evaluate if the student's solution is correct or not.
Don't decide if the student's solution is correct until
you have done the problem yourself.

Use the following format:
Question:

question here

Student's solution:

student’s solution here

Actual solution:

steps to work out the solution and your solution here

Is the student's solution the same as actual solution \
just calculated:

yes or no

Student grade:

correct or incorrect

Question:

I’m building a solar power installation and I need help
working out the financials.

• Land costs $100 / square foot
• I can buy solar panels for $250 / square foot
• I negotiated a contract for maintenance that will cost
  me a flat $100k per year, and an additional $10 / square
  foot What is the total cost for the first year of operations
  as a function of the number of square feet.

Student's solution:

Let x be the size of the installation in square feet. Costs:

1. Land cost: 100x
2. Solar panel cost: 250x
3. Maintenance cost: 100,000 + 100x Total cost: 100x + 250x + 100,000 + 100x = 450x + 100,000

Actual solution:
"""
response = get_completion(prompt)
print(response)

写作指南

简单代码

格式：Write a [describe the tech stack] app using [any auxiliary tech like a database]. I need it to [describe the most important requirements].

例子：Write a Python/FastAPI app using a PostgreSQL database, hooked up with SQLAlchemy. I need to accept HTML requests and write them to the database

复杂代码

例子：You are an expert backend developer. Your task is to write a Python/FastAPI app with a PostgreSQL database. The application should accept HTML requests and write them to the database.Write a detailed tech spec, including user requirements for [your user type goes here]

提高问题解决能力

格式：Share a step-by-step systematic approach for solving [specific problem or challenge].

改进文本

使用 LLM 来对自己的文本进行改进，尽量不要使用翻译功能，因为用原语言书写也能提高自己的写作能力。

格式：[Paste Your Writing]. Proofread the above text for spelling and grammar. Make the sentences more clear and native.

逆向工程

如果不知道怎样使用 Prompt 向 ChatGPT 提出问题，只需告诉 ChatGPT 即可。

What’s the best prompt for ChatGPT to learn my writing styles and respond to my emails for me?