Keras – 2. Software & Hardware Specification

Welcome to the CS With James

In this post I will explain the softwares and the hardware I will going to use

I will use Python3 as the default programming language, but python2 shares lots of common grammars with the python3 so, python2 would not be a problem.

I will use Tensorflow and the Keras as a Machine Learning library. I believe Keras also support CNTK and Theano. If you are more familiar with CNTK and Theano feel free to use those libraries, but I do not guarantee all the code will run smoothly just doing the copy&paste from my code.

I will use MacOS as and Mac computer as my developing environment. However, as long as you have all the softwares installed in your computer there should not have problem following my tutorials using Windows and Unix/Linux. If you have Windows machine, then I highly suggest you to install a Ubuntu in your computer while you learning the Machine Learning. I saw lots of people had issues with running Tensorflow on Windows.

I will use Nvidia’s graphic card because Machine Learning libraries only support CUDA which is running on the Nvidia’s card only. Unfortunately the AMD Card doesn’t support the GPU acceleration at this time. However, you can use CPU only version of the TensorFlow to learn the Machine Learning. If you want to go deeper in the Machine Learning I highly suggest to get a Nvidia’s card.

For more information about what GPU you should get for the Machine Learning please click here for more information.

Leave a Reply

Your email address will not be published. Required fields are marked *