Using conversational memory in Tool used by agent

[cvgoudar]

I have custom tool which internally uses a LLM Chain. However I want to use conversational memory within the tool. How to pass the memory to tool. I would not like to have the conversational memory as global.

I passed the memory in Tool.from_function call and it threw error that memory argument is invalid.

[dosubot[bot]]

🤖

Hello,

To incorporate conversational memory into your custom tool that uses a LLM Chain without making it global, you can use the ZepMemory class provided by the LangChain framework. This class extends the ConversationBufferMemory and persists your chain history to the Zep Memory Server, which is not global and is specific to each instance of the ZepMemory.

Here is an example of how you can use it:

from langchain.chains import ConversationChain
from langchain.llms import OpenAI
from langchain.memory.zep_memory import ZepMemory

# Create a new instance of ZepMemory
memory = ZepMemory(
    session_id=session_id,  # Identifies your user or a user's session
    url=ZEP_API_URL,        # Your Zep server's URL
    api_key=<your_api_key>, # Optional
    memory_key="history",   # Ensure this matches the key used in chain's prompt template
    return_messages=True,   # Does your prompt template expect a string or a list of Messages?
)

# Create a new instance of ConversationChain with OpenAI as the language model and ZepMemory as the memory store
conversation = ConversationChain(llm=OpenAI(), memory=memory)

In this example, conversation is an instance of ConversationChain that uses OpenAI as its language model and ZepMemory as its memory store. It has its own ZepMemory instance as its memory store, which is not shared with other instances of ConversationChain.

The ZepMemory class also provides a save_context method that saves the context from this conversation to the buffer. This helps to persist metadata alongside your chat history.

You can find more details about the ZepMemory class in the LangChain framework in the following files:

• libs/langchain/langchain/chains/conversation/base.py
• libs/langchain/langchain/memory/zep_memory.py

I hope this helps! If you have any other questions, feel free to ask.

Best,
Dosu

Sources

• libs/langchain/langchain/chains/conversation/base.py
• libs/langchain/langchain/memory/zep_memory.py
• libs/langchain/tests/integration_tests/memory/chat_message_histories/test_zep.py

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.