This repo contains a from-scratch Implementation of a Transformer Model in tensorflow/keras, following the original paper "Attention is all you need", in model.py. The model is applied in the recommendation-domain to predict products from a sequence.