五、智能体架构:Agent
5.1 什么是智能体(Agent)
将大语言模型作为一个推理引擎。给定一个任务,智能体自动生成完成任务所需的步骤,执行相应动作(例如选择并调用工具),直到任务完成。
5.2 先定义一些工具:Tools
- 可以是一个函数或三方 API
- 也可以把一个 Chain 或者 Agent 的 run()作为一个 Tool
from langchain import SerpAPIWrapper
search = SerpAPIWrapper()
tools = [
Tool.from_function(
func=search.run,
name="Search",
description="useful for when you need to answer questions about current events"
),
]
from langchain.tools import Tool, tool
import calendar
import dateutil.parser as parser
from datetime import date
@tool("weekday")
def weekday(date_str: str) -> str:
"""Convert date to weekday name"""
d = parser.parse(date_str)
return calendar.day_name[d.weekday()]
from langchain.agents import load_tools
tools = load_tools(["serpapi"])
tools += [weekday]
5.3 智能体类型:ReAct
!pip install google-search-results
from langchain.chat_models import ChatOpenAI
from langchain.llms import OpenAI
from langchain.agents import AgentType
from langchain.agents import initialize_agent
llm = ChatOpenAI(model_name=‘gpt-4’, temperature=0)
agent = initialize_agent(
tools, llm, agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION, verbose=True)
agent.run(“周杰伦生日那天是星期几”)
5.4 通过 OpenAI Function Calling 实现智能体
from langchain.chat_models import ChatOpenAI
from langchain.llms import OpenAI
from langchain.agents import AgentType
from langchain.agents import initialize_agent
llm = ChatOpenAI(model_name=‘gpt-4-0613’, temperature=0)
agent = initialize_agent(
tools,
llm,
agent=AgentType.OPENAI_FUNCTIONS,
verbose=True,
max_iterations=2,
early_stopping_method=“generate”,
)
agent.run(“周杰伦生日那天是星期几”)
5.5 智能体类型:SelfAskWithSearch
from langchain import OpenAI, SerpAPIWrapper
from langchain.agents import initialize_agent, Tool
from langchain.agents import AgentType
llm = OpenAI(temperature=0)
search = SerpAPIWrapper()
tools = [
Tool(
name=“Intermediate Answer”,
func=search.run,
description=“useful for when you need to ask with search”,
)
]
self_ask_with_search = initialize_agent(
tools, llm, agent=AgentType.SELF_ASK_WITH_SEARCH, verbose=True
)
self_ask_with_search.run(
“冯小刚的老婆演过什么电影”
)
5.6 智能体类型:Plan-and-Execute
!pip install langchain-experimental
from langchain.utilities.wolfram_alpha import WolframAlphaAPIWrapper
from langchain.agents import load_tools
from langchain import SerpAPIWrapper
from langchain.agents.tools import Tool
from langchain.llms import OpenAI
from langchain_experimental.plan_and_execute import PlanAndExecute, load_agent_executor, load_chat_planner
from langchain.chat_models import ChatOpenAI
from langchain.memory import ConversationSummaryMemory
llm = ChatOpenAI(model_name=‘gpt-4’, temperature=0)
search = SerpAPIWrapper(params={
‘engine’: ‘google’,
‘gl’: ‘cn’,
‘google_domain’: ‘google.com.hk’,
‘hl’: ‘zh-cn’
})
tools = [
Tool(
name=“Search”,
func=search.run,
description=“useful for when you need to answer questions about current events”
)
]
planner = load_chat_planner(llm)
executor = load_agent_executor(llm, tools, verbose=True)
agent = PlanAndExecute(planner=planner, executor=executor, verbose=True)
agent.run(“分析北京明天天气,与上海明天天气对比,用中文写一遍报告”)
后记
📢博客主页:https://manor.blog.csdn.net文章来源:https://www.toymoban.com/news/detail-813788.html
📢欢迎点赞 👍 收藏 ⭐留言 📝 如有错误敬请指正!
📢本文由 Maynor 原创,首发于 CSDN博客🙉
📢不能老盯着手机屏幕,要不时地抬起头,看看老板的位置⭐
📢专栏持续更新,欢迎订阅:https://blog.csdn.net/xianyu120/category_12471942.html文章来源地址https://www.toymoban.com/news/detail-813788.html
到了这里,关于AI全栈大模型工程师(十六)智能体架构:Agent的文章就介绍完了。如果您还想了解更多内容,请在右上角搜索TOY模板网以前的文章或继续浏览下面的相关文章,希望大家以后多多支持TOY模板网!