This notebook has evolved throughout our experimentation, and cannot be seen as a linear history of our ork, and cannot be run through from one end to the other. However all the methods we used to train our models are here. A model can be transfer learned in the transfer learning section, and can be fine tuned in the fine tuning section. After training the desired model one should can compute and print the performance statistics in the Model Results section and create many different visualsation of the model performance in the rest of the sections.