# TensorFlow Tutorial - Pretty Tensor

Notes by Magnus Erik Hvass Pedersen: https://github.com/Hvass-Labs/TensorFlow-Tutorials/blob/master/03_Pretty...

The previous tutorial showed how to implement a Convolutional Neural Network in TensorFlow, which required low-level knowledge of how TensorFlow works. It was complicated and easy to make mistakes.

This tutorial shows how to use the add-on package for TensorFlow called PrettyTensor, which is also developed by Google. PrettyTensor provides much simpler ways of constructing Neural Networks in TensorFlow, thus allowing us to focus on the idea we wish to implement and not worry so much about low-level implementation details. This also makes the source-code much shorter and easier to read and modify.

Most of the source-code in this tutorial is identical to Tutorial #02 except for the graph-construction which is now done using PrettyTensor, as well as some other minor changes.

This tutorial builds directly on Tutorial #02 and it is recommended that you study that tutorial first if you are new to TensorFlow. You should also be familiar with basic linear algebra, Python and the Jupyter Notebook editor.