This repository contains the code necessary to train and evaluate a model using TFlearn. I trained 8 different neural networks on the same dataset to compare how they behave.

First create a npy format testset and a trainingset using the create_testset.py create_trainingset.py scripts.
The next step is to train the 8 models using the train.py file. There is a variant to train the same models using
data augmentation techniques.
The final step is to evaluate the models using one of the first 3 scripts.