[OpenAI API-4] Prompting Techniques

파이썬으로 도메인 특화 챗봇 모델 만들기 - 4

 

Open Ai git에 open ai의 신뢰성을 높일 수 있는 다양한 기술들이 기재되어 있다.

그 중 프롬프트 엔지니어링 부분에서 인상 깊었던 내용들을 정리해보고자 한다.

 

- 목 차 -

1. Text completion vs Chat completion
2. 프롬프트 엔지니어링 테크닉 1) Chain of Thought Prompting
3. 프롬프트 엔지니어링 테크닉 2) Least-to-most-Prompting

 

---

 

Text completion vs Chat completion

2개가 어떻게 다른지 차이점을 잠깐 보자면 

 

Text completion

• prompt에 모든 내용을 집약해놓은 형태
• 따라서, 프롬프트 엔지니어링은 저 프롬프트에 넣을 문장을 엔지니어링 해야한다.

import os
import openai

openai.api_key = os.getenv("OPENAI_API_KEY")

response = openai.Completion.create(
  model="text-davinci-003",
  prompt="I am a highly intelligent question answering bot. If you ask me a question that is rooted in truth, I will give you the answer. If you ask me a question that is nonsense, trickery, or has no clear answer, I will respond with \"Unknown\".\n\nQ: What is human life expectancy in the United States?\nA: Human life expectancy in the United States is 78 years.\n\nQ: Who was president of the United States in 1955?\nA: Dwight D. Eisenhower was president of the United States in 1955.\n\nQ: Which party did he belong to?\nA: He belonged to the Republican Party.\n\nQ: What is the square root of banana?\nA: Unknown\n\nQ: How does a telescope work?\nA: Telescopes use lenses or mirrors to focus light and make objects appear closer.\n\nQ: Where were the 1992 Olympics held?\nA: The 1992 Olympics were held in Barcelona, Spain.\n\nQ: How many squigs are in a bonk?\nA: Unknown\n\nQ: Where is the Valley of Kings?\nA:",
  temperature=0,
  max_tokens=100,
  top_p=1,
  frequency_penalty=0.0,
  presence_penalty=0.0,
  stop=["\n"]
)

 

Chat completion

• model : gpt-3.5-turbo, gpt-4 사용 가능
• messages
  • system : 챗봇의 역할 부여, 어시스턴트의 동작을 설
  • user : 질문 형식
  • assistant : 대답 형식 등

# Note: you need to be using OpenAI Python v0.27.0 for the code below to work
import openai

openai.ChatCompletion.create(
  model="gpt-3.5-turbo",
  messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Who won the world series in 2020?"},
        {"role": "assistant", "content": "The Los Angeles Dodgers won the World Series in 2020."},
        {"role": "user", "content": "Where was it played?"}
    ]
)

 

프롬프트 엔지니어링 테크닉

Chain of Thought Prompting

: 계산과정을 추가하여 줌으로써, 유사한 형식의 질문에 대해서 명확한 대답을 얻을 수 있음.

Chain of Thought Prompting -1

Chain of Thought Prompting -2 QA 특화

 

Least-to-most prompting

: 하위 작업을 분할하는 방법

• 예)  “워터슬라이드 몇번 탈수 있냐”의 질문에 대한 답을 얻고 싶을때, 그 전 계산되어야 할것들을 서브 퀘스쳔&대답을 추가하는 방법

 

참고 )

openai-cookbook/techniques_to_improve_reliability.md at main · openai/openai-cookbook

Least-to-Most Prompting Enables Complex Reasoning in Large Language Models