2024년 10월 6일 일요일

LangGraph 기반 데이터 분석 멀티 에이전트 만들기

이 글은 LangGraph 기반 데이터 분석 멀티 에이전트 만드는 방법을 간략히 설명한다.

LangChain Tool 기반 수학 에이전트 개발
다음 코드를 입력해 실행한다. 

from langchain_openai import OpenAI
from langchain_community.chat_models import ChatOllama
from langchain.chains import LLMMathChain, LLMChain
from langchain.prompts import PromptTemplate
from langchain_community.utilities import WikipediaAPIWrapper
from langchain.agents.agent_types import AgentType
from langchain.agents import Tool, initialize_agent
from dotenv import load_dotenv
import chainlit as cl

load_dotenv()

@cl.on_chat_start
def math_chatbot():
llm = ChatOllama(model='llama3', temperature=0.0)

  # prompt for reasoning based tool
word_problem_template = """You are a reasoning agent tasked with solving t he user's logic-based questions. Logically arrive at the solution, and be factual. In your answers, clearly detail the steps involved and give the final answer. Provide the response in bullet points. Question  {question} Answer"""

math_assistant_prompt = PromptTemplate(
input_variables=["question"],
template=word_problem_template
)
  # chain for reasoning based tool
word_problem_chain = LLMChain(llm=llm,
  prompt=math_assistant_prompt)
# reasoning based tool                              
word_problem_tool = Tool.from_function(name="Reasoning Tool",
   func=word_problem_chain.run,
   description="Useful for when you need to answer logic-based/reasoning questions."
   )
  # calculator tool for arithmetics
problem_chain = LLMMathChain.from_llm(llm=llm)
math_tool = Tool.from_function(name="Calculator",
   func=problem_chain.run,
   description="Useful for when you need to answer numeric questions. This tool is only for math questions and nothing else. Only input math expressions, without text",
   )
  
  # Wikipedia Tool
wikipedia = WikipediaAPIWrapper()
wikipedia_tool = Tool(
name="Wikipedia",
func=wikipedia.run,
description="A useful tool for searching the Internet to find information on world events, issues, dates, "
"years, etc. Worth using for general topics. Use precise questions.",
)
  
  # agent
agent = initialize_agent(
tools=[wikipedia_tool, math_tool, word_problem_tool],
llm=llm,
agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION,
verbose=False,
handle_parsing_errors=True
)
cl.user_session.set("agent", agent)

@cl.on_message
async def process_user_query(message: cl.Message):
agent = cl.user_session.get("agent")

response = await agent.acall(message.content,
callbacks=[cl.AsyncLangchainCallbackHandler()])

await cl.Message(response["output"]).send()

결과는 다음과 같다.


레퍼런스

작성자: 시간:

댓글 없음:

댓글 쓰기

피드 구독하기: 댓글 (Atom)