Passing variable to multi-turn conversation eval prompt, coming as empty during execution

LangSmith Product Help Observability & Evals
1

I’m trying to predict the sentiment of a thread. But while passing the thread history (conversation) to the prompt, it displays [‘all_message_objects’, ‘all_messages’, ‘first_human_last_ai’, ‘human_ai_pairs’] but all variables are coming as empty while executing.

image
image1920×1440 193 KB

2

hi @guna-20

what’s your code? Are you running it via RunnableSequence? Share your code please.

3

@pawel-twardziak

this is the python code to execute the agent

          executor = AgentExecutor(agent=self.agent,                tools=self.tools,

            verbose=False,

            memory=memory,

            return_intermediate_steps=True,

            handle_parsing_errors=True,

        )

        config: RunnableConfig = {“tags”: [f"brand:{self.brand}"],“metadata”: {“thread_id”: str(sender_id), “sender_id”: sender_id,“brand”: self.brand,            }}

        start = time.time()



        result = await executor.ainvoke(

            input={“input”: input,“sender_id”: sender_id,“ticket_id”: str(ticket_id),           },

            config=config,

        )

and it is adding the threads in langsmith like this

image
image1738×957 140 KB

i m using default multi-turn conversation eval in the langsmith dashboard to predict the sentiment of the user. During the execution of the multi-turn eval the messages are coming as empty string