0

I'm using langchain python because I'm working on creating a custom knowledge chatbot. I have created the vecstores and everything works fine until I introduced LangChain's agents. The CHAT_CONVERSATIONAL_REACT_DESCRIPTION works well but sometimes I get this error when asking query: "json.decoder.JSONDecodeError: Unterminated string starting at: line 3 column 21 (char 52)"

This is an example of what happens

`

Ask your question: suggest 2 vacation locations
response:


> Entering new AgentExecutor chain...
Traceback (most recent call last):
  File "C:\Users\qais4\Desktop\chatbot test\tutor.py", line 372, in <module>
    run_terminal()
  File "C:\Users\qais4\Desktop\chatbot test\tutor.py", line 365, in run_terminal
    get_response(agent, question)
  File "C:\Users\qais4\Desktop\chatbot test\tutor.py", line 206, in get_response
    result = agent.run(question)
             ^^^^^^^^^^^^^^^^^^^
  File "C:\Users\qais4\Desktop\chatbot test\venv\Lib\site-packages\langchain\chains\base.py", line 236, in run
    return self(args[0], callbacks=callbacks)[self.output_keys[0]]
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\qais4\Desktop\chatbot test\venv\Lib\site-packages\langchain\chains\base.py", line 140, in __call__
    raise e

This is how I set up the agent:

 agent = initialize_agent(
        agent=AgentType.CHAT_CONVERSATIONAL_REACT_DESCRIPTION,
        tools=tools,
        llm=ChatLLM,
        verbose=True,
        memory=conversational_memory,
    )

Does anyone have any suggestions?

I pretty much tried all agent types but all of them get errors on some specific case. The only one without errors was CONVERSATIONAL_REACT_DESCRIPTION Agent type but that often makes bad decisions in terms of tools (choosing not to select knowledge base tool and saying "I dont know"). I have also tried following the examples on Langchain's documentation but that didn't work.

user21966626
  • 1

0 Answers0