[GPT][INVESTORGPT] Zero-shot ReAct Agent

 

from langchain.chat_models import ChatOpenAI

from langchain.tools import Tool

from langchain.agents import initialize_agent, AgentType

llm = ChatOpenAI(

    temperature=0.1,

)

def plus(inputs):

    a,b = inputs.split(",")

    return float(a) + float(b)

agent = initialize_agent(

    llm=llm,

    verbose=True,

    agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION,

    handle_parsing_erros=True,

    tools=[

        Tool.from_function(

            func=plus,

            name="Sum Calculator",

            description="Use this to perform sums of two numbers. This tool take two arguments, both should be numbers. till finish. Use this tool by sending a pair of number separated by a comma. \n Example: 2,3",

        )

    ],

)

prompt = "Tell me Total Cost of $355.39 + $924.87 + $721.2 + $1940.29 + $573.63 + $65.72 + $35.00 + $552.00 + $76.16 + $29.12"

#llm.invoke(prompt)

agent.invoke(prompt)

 

Zero-shot ReAct

 

작업과 도구 설명을 보고 사용할 도구를 결정

ReAct는 Reasoning + Acting으로 Yao et al. (2022)의 논문에 소개된 개념

Reasoning은 생각(Chain-of-Thought, COT), Acting은 외부 도구 실행을 의미

필요한 작업에 따라 정하면 되는데, Zero-shot ReAct가 가장 기본이 되는 Agent

 

전달 받는값은 1개로 제한되며, 이를 분리하여 사용한다.