Are you working on Deep Learning/Machine Learning/AI based projects? If so, this tutorial is for you! In this tutorial, we go over how to setup TensorFlow on a VPS or Dedicated Server.

Due to higher resource/CPU requirements, we highly recommend running TensorFlow within a dedicated server (instead of a VPS) environment.

Before deep diving into this tutorial – let’s first go over the prerequisites.

Prerequisites:

The prerequisites remain the same whether you are running a Linux Distribution or Windows OS:

1. You VPS instance should have a memory of at least for 4GB for working smoothly
2. If you’re utilizing Ubuntu Linux, please note that TensorFlow is supported on version 16.04 and above.

Steps to Install TensorFlow on Ubuntu

Setup and install Python development on your system

If you don’t have python or pip or virtual environment installed you can install using below steps

sudo apt-get update -y

sudo apt install python3-dev python3-pip -y

sudo pip3 install -U virtualenv

Update Pip

python3 -m pip install –upgrade pip

Install TensorFlow

pip install –user –upgrade tensorflow

Validate Installation

You might see some warning messages libcudart, you can ignore those since you’re running it on server environments, the warning is caused because some GPU based library could not be loaded.

python3

>> import tensorflow as tf

>> print (tf.__version__)

If you are running a Windows OS, the same steps above are mostly applicable. You will simply need to download python and pip from official page, and then proceed with further steps given in this tutorial.

If you are looking for a TensorFlow Dedicated Server (a dedicated server you can run TensorFlow on), be sure to take a look at RackNerd Dedicated Servers.