[P] My deep learning library from sophomore year in high school

I made a deep learning library in Java a while ago to help me understand how neural networks work. I implemented vanilla fully connected layers, convolutional layers, pooling layers, GRU cells, multiple gradient descent optimizers, and other cool stuff. I also strived for an object oriented framework with a Keras-like interface.

It took me quite a while to understand the full backprop algorithm as I did not know how derivatives work when I started. Also, many intrinsic details about how neural networks operate were obscured behind very similar mathematical equations in most of the blog posts I've seen (although Stanford CS231 was very helpful). In the end, I derived the backprop equations for updating the weights by hand for all of the layers in the library. As such, I want to share my code and maybe help others that are learning about neural networks.

Note the library is very slow. There are no fancy GPU speedups because the library is pure Java. You can test the pretrained MNIST by running TestMNISTDraw2.java in the src/tests folder that also contains other demos.

Edit: Is this too "beginner" for this sub?