Exploration of Deep Learning pipelines made easy

The easiest way to show how ATOM can help you is through an example. This story walks you through a notebook that trains and validates a Convolutional Neural Network implemented with Keras. The model is trained using MNIST¹, a well known image dataset whose goal is to classify handwritten digits. We start with the necessary imports and defining the model: a simple neural network with two convolutional layers and one output layers consisting of 10 neurons, one for each digit. The dataset contains 28x28 grayscale images, therefore every image's array has shape (28, 28, 1).