feat: add tool calling for chatgpt-based math agent

[dongyuanjushi]

1. add tool calling for chatgpt-based math agent
2. refactor and wrap the LLM request with the Message object and wrap the LLM response with the Response object

[TataKKKL]

could you log more information on function calling, consider these information

function_name = tool_call.function.name
function_to_call = available_functions[function_name]
function_args = json.loads(tool_call.function.arguments)
function_response = function_to_call(**function_args)

messages.append(
    {
        "tool_call_id": tool_call.id,
        "role": "tool",
        "name": function_name,
        "content": str(function_response),
    }

[TataKKKL]

Currently, you are modifying the output of a function call using natural language and some heuristics. However, ChatGPT has a workaround below, which will need to be modified later.

for tool_call in tool_calls:
    function_name = tool_call.function.name
    function_to_call = available_functions[function_name]
    function_args = json.loads(tool_call.function.arguments)
    function_response = function_to_call(**function_args)
  
    messages.append(
        {
            "tool_call_id": tool_call.id,
            "role": "tool",
            "name": function_name,
            "content": str(function_response),
        }
    )  # extend conversation with function response
second_response = client.chat.completions.create(
  model="gpt-4-1106-preview",
  messages=messages,
)  # get a new response from the model where it can see the function response