In streaming mode, how would you add final reponse to chat history?

[longpk]

I'm having this chatbot with function calling in streaming mode as follows. I don't know how to correctly add the final responese to the chat history.

userInput = None
while True:
    userInput = input("User > ")

    if userInput == "exit":
        break

    history.add_user_message(userInput)

    response = chat_completion.get_streaming_chat_message_contents(
        chat_history=history,
        settings=execution_settings,
        kernel=kernel,
        arguments=KernelArguments(),
    )

    result = None
    print("Assistant > ", end="")
    async for chunk in response:
        print(str(chunk[0]), end="")
        if result is None:
            result = chunk[0]
        else:
            result = result + chunk[0]
    print("")

    # Don't know how to add response to history. The following line would raise exception in the next chat turn.
    history.add_message(result)

The code will cause this exception:

semantic_kernel.exceptions.service_exceptions.ServiceResponseException: ("<class 'semantic_kernel.connectors.ai.open_ai.services.azure_chat_completion.AzureChatCompletion'> service failed to complete the prompt", BadRequestError('Error code: 400 - {\'error\': {\'message\': "An assistant message with \'tool_calls\' must be followed by tool messages responding to each \'tool_call_id\'. The following tool_call_ids did not have response messages: call_pAvHKESwoWEMDDyNfitPu0Ja, call_0V4KTml2pfM1sPPsYIsvg9N0", \'type\': \'invalid_request_error\', \'param\': \'messages.[9].role\', \'code\': None}}'))

[akordowski]

In C# there is an Metadata property on the StreamingChatMessageContent class. The last message contains Metadata["FinishReason"] = "stop" with which you can check if the message is the last. I hope you can find something similar in Python.

[moonbox3]

Hi @longpk, in Python, we have the following auto function calling concept sample script that shows how to handle streaming messages, and adding them to the chat history when done. You can see in the async def handle_streaming(...) method, that we build up the result_content list, if applicable, and then join those items once the method is done.

semantic-kernel/python/samples/concepts/auto_function_calling/chat_gpt_api_function_calling.py

Line 120 in 9d4966a

async def handle_streaming(

[longpk]

Thank you for pointing to the right direction. Looking at sample script, I changed this line:
history.add_message(result)
to
history.add_assistant_message(result)

and it works.
Thanks for your help.