After #16 , integrate VQA with the identified model serving backend. in the VQA page, I want to be able to select any model that is supported and inference with that model with vision request (image and text)
