增加memory的方法
给langchain的内置agent增加memory，本方法首先通过 ConversationBufferMemory 实例化并传递到 initialize_agent 的 memory 参数中，从而实现对话记忆的功能，还使用MessagePlaceHolder，该方法可以将memory的key传递到提示词模版中，提升agent记忆能力。
from langchain.utilities import SerpAPIWrapper from langchain.chains import LLMMathChain from langchain.agents import initialize_agent, AgentType import os from langchain.agents import Tool,load_tools from langchain.memory import ConversationBufferMemory from langchain.prompts import MessagesPlaceholder # serppai的token os.environ["SERPAPI_API_KEY"] = "95ac0e518f8e578cc81b149144efd7535d5d7ccab87244e946a1cf3bb14ef3e7" class AgentsTemplate:def __init__(self,**kwargs):#构建一个搜索工具search = SerpAPIWrapper()self.prompt = kwargs.get("base_prompt")self.llm = kwargs.get("llm")llm_math_chain = load_tools(["serpapi", "llm-math"], llm=self.llm)self.tools = [Tool(name="Search",func=search.run,description= "useful for when you need to answer questions about current events or the current state of the world"),Tool(name="Math Chain",func=llm_math_chain[1].run,description="useful for solving mathematical problems")]#load_tools(["serpapi", "llm-math"], llm=self.llm)# 记忆组件self.memory = ConversationBufferMemory(memory_key="chat_history",return_messages=True,)self.agentType = [AgentType.ZERO_SHOT_REACT_DESCRIPTION,AgentType.CHAT_ZERO_SHOT_REACT_DESCRIPTION,AgentType.CONVERSATIONAL_REACT_DESCRIPTION,AgentType.CHAT_CONVERSATIONAL_REACT_DESCRIPTION,AgentType.STRUCTURED_CHAT_ZERO_SHOT_REACT_DESCRIPTION]#零样本增强式生成ZERO_SHOT_REACT_DESCRIPTION,#使用chatModel的零样本增强式生成CHAT_ZERO_SHOT_REACT_DESCRIPTION,def zero_agent(self,question,agentType):if agentType not in self.agentType:raise ValueError("无效的 AgentType，请选择有效的类型！")# 动态构建初始化参数agent_params = {"tools": self.tools,"llm": self.llm,"agent": agentType,"verbose": True,"memory": self.memory,"agent_kwargs" : {"extra_prompt_messages": [MessagesPlaceholder(variable_name="chat_history"),MessagesPlaceholder(variable_name="agent_scratchpad")],},"handle_parsing_errors": True}#初始化代理agent = initialize_agent(**agent_params)print("-------------------")# 输出提示词模板prompt = agent.agent.llm_chain.promptprint("Prompt Template:")print(prompt)# print(agent.agent.prompt.messages)# print(agent.agent.prompt.messages[0])# print(agent.agent.prompt.messages[1])# print(agent.agent.prompt.messages[2])try:response = agent.run(question)print(f"运行的代理类型: {agentType}, 提问内容: {question}")print(f"agent回答: {response}")#self.memory.save_context(question,response)except Exception as e:print(f"代理运行时出错: {e}")#使用chatModel的零样本增强式生成
其他添加memory的方法
LangChain 提供了多种 Memory 类型
1. ConversationBufferWindowMemory: 仅存储最近的 k 条消息，适合避免内存无限增长。
2. ConversationSummaryMemory: 将对话总结成摘要以保留上下文。
3. CombinedMemory: 可以组合多种 Memory 类型。
动态保存和加载 Memory: 该方法可以在对话结束时保存 memory 并在新对话时重新加载：
# 保存 memory with open("memory.json", "w") as file:file.write(self.memory.json())# 加载 memory from langchain.memory import ConversationBufferMemory with open("memory.json", "r") as file:memory_data = file.read() self.memory = ConversationBufferMemory.from_json(memory_data)
