We will be feeding data to a generic chatbot class and calling the vectorize methods inside it. We will use it as a wrapper for the memory network model we defined earlier. In theory, the model can be swapped out for any other QA-based model.