Created by: LielinJiang

Add vision api such as transforms(Compose, RandomResizedCrop, Normalize...) and vision models api(resnet, mobilenet, vgg...).

tree docs: