rasbt/deeplearning-models ? reverse-engineered prompt

Build me a clean GitHub repo of Jupyter notebooks that teaches deep learning by example, using both PyTorch and TensorFlow.

I want a well organized collection of notebooks for the basics like perceptron, logistic regression, softmax regression, then multilayer perceptrons with things like dropout, batch normalization, and a from scratch backprop version, plus a solid set of CNN examples for computer vision like LeNet, AlexNet, DenseNet, MobileNet, and a few concept notebooks such as replacing fully connected layers with convolutions. Keep it beginner friendly, with short explanations, runnable code, and examples on common datasets like MNIST, CIFAR 10, Iris, and toy data where that makes sense.

Please make the README easy to scan with a table of notebooks and links, and keep the folders tidy so it’s simple to find the PyTorch, PyTorch Lightning, and TensorFlow versions. If you need to check current notebook syntax or library docs online, go ahead and do that.