题意：如何持久化 LangChain 对话记忆（保存和加载）

问题背景：

I'm creating a conversation like so:        我正在创建一个对话，如下所示：

llm = ChatOpenAI(temperature=0, openai_api_key=OPENAI_API_KEY, model_name=OPENAI_DEFAULT_MODEL)
conversation = ConversationChain(llm=llm, memory=ConversationBufferMemory())

But what I really want is to be able to save and load that ConversationBufferMemory() so that it's persistent between sessions. There doesn't seem to be any obvious tutorials for this but I noticed "Pydantic" so I tried to do this:

但我真正想要的是能够保存和加载 ConversationBufferMemory()，以便它在会话之间保持持久性。似乎没有关于这方面的明显教程，但我注意到了“Pydantic”，所以我尝试这样做：

saved_dict = conversation.memory.chat_memory.dict()
cm = ChatMessageHistory(**saved_dict) # or cm = ChatMessageHistory.parse_obj(saved_dict)

But this fails:        但是失败了

ValidationError: 6 validation errors for ChatMessageHistory
messages -> 0Can't instantiate abstract class BaseMessage with abstract method type (type=type_error)

Thoughts? I'd love links to any sort of guide, repo, reference, etc.

想法？我很乐意看到任何相关的指南、仓库、参考等链接。

问题解决：

I just did something similar, hopefully this will be helpful. On a high level:

我刚刚做了一些类似的事情，希望能有所帮助。在高层次上：

1. use ConversationBufferMemory as the memory to pass to the Chain initialization

使用 ConversationBufferMemory 作为内存传递给链（Chain）的初始化

llm = ChatOpenAI(temperature=0, model_name='gpt-3.5-turbo-0301')
original_chain = ConversationChain(llm=llm,verbose=True,memory=ConversationBufferMemory()
)
original_chain.run('what do you know about Python in less than 10 words')

2. extract messages from memory in the form of List[langchain.schema.HumanMessage|AIMessage] (not serializable)

从内存中提取消息，以 List[langchain.schema.HumanMessage | AIMessage] 的形式（不可序列化）

extracted_messages = original_chain.memory.chat_memory.messages

3. transform the extracted message to serializable native Python objects

将提取的消息转换为可序列化的原生Python对象

ingest_to_db = messages_to_dict(extracted_messages)

4. perform db operations to write to and read from database of your choice, I'll just use json.dumps and json.loads to illustrate

执行数据库操作以写入和读取您选择的数据库，为了说明，我将只使用 json.dumps 和 json.loads

retrieve_from_db = json.loads(json.dumps(ingest_to_db))

5. transform the retrieved serialized object back to List[langchain.schema.HumanMessage|AIMessage]

将检索到的序列化对象转换回 List[langchain.schema.HumanMessage | AIMessage] 的形式

retrieved_messages = messages_from_dict(retrieve_from_db)

6. construct ChatMessageHistory from the messages

从消息中构建 ChatMessageHistory

retrieved_chat_history = ChatMessageHistory(messages=retrieved_messages)

7. construct ConversationBufferMemory from ChatMessageHistory

从 ChatMessageHistory 中构建 ConversationBufferMemory

retrieved_memory = ConversationBufferMemory(chat_memory=retrieved_chat_history)

8. pass memory back to the newly initiated Chain

将内存（memory）传递回新初始化的链（Chain）

reloaded_chain = ConversationChain(llm=llm,verbose=True,memory=retrieved_memory
)