given dataset of images that contains $10$ M images and $100$ classes, draw VAE architechure for learning from this datset. EXplain the difference between encoder and VAE, the goal of each? what the input and the label for network