torchnets
This is just a small project that I started to understand more about the construction of neural networks, particularly for computer vision. TL;DR: it helped a lot.
This python-installable package runs on pytorch (I try to keep it compatible with the latest version) and implements a variety of famous neural nets from recent deep learning history. For each model, I try to cite the relevant paper that introduced it, and with any luck this will be a nice archive of neural net examples.
DISCLAIMER This package is a personal experiment. Use the networks within at your own risk. These networks are not distributed with pre-trained weights.