Gemini GenerativeModel.start_chat() is not supported

[syberkitten]

Is your feature request related to a problem? Please describe.
Using the start_chat mechanism with instructor for gemini does not work.
documentation shows example with chat.completions.create, but I'm using start_chat
allows more flexibility with supplemental files.
start_chat() returns a chat session object on which we run:
chat.send_message_async(question, stream, safety_settings)
adding response_model to it fails:
response = await self.chat_session.send_message_async(question, stream=False, TypeError: ChatSession.send_message_async() got an unexpected keyword argument 'response_model'

Describe the solution you'd like
support the response_model in send_message_async

Describe alternatives you've considered
not using gemini with instructor at this point

Additional context
instructor = "^1.6.3"
google-generativeai = "^0.8.2"