Is your feature request related to a problem? Please describe.
I cannot customize the model configs like temperature or top_p when creating a Chat, which prevents me from achieving better results with the Chat's generate_message function.

Describe the solution you'd like
The inference method of the Session in app.services.assistant.generation.session should not hardcode the configs as {}, but instead, it should accept the configs set by the Chat. Correspondingly, the Chat class should also include a configs field, and the related interfaces of the Python client should be modified accordingly.

Describe alternatives you've considered
Currently, I can only set configs using taskingai.inference.chat_completion. At the meantime, to maintain chat functionality, I would need to manually manage the history of messages.

Additional context
Add any other context or screenshots about the feature request here.