DeepLearn

Why we included this project

Reading a paper and wondering what the model actually looks like in code is a common frustration, and this repo is built around exactly that gap. It collects runnable Python implementations of well-known deep learning papers, each one linking to the original PDF and a folder of Keras or TensorFlow code: attention-based CNNs for sentence pairs, siamese recurrent networks for similarity, neural tensor networks for knowledge-base completion. The focus leans heavily toward NLP, with question answering and stance detection well represented, though representation learning and correlation-based computer vision get space as well. It isn't a maintained library for production work, so think of it as a study aid for moving from theory to a working model.