Introduction

TensorFlow is a free and open-source software library for machine learning and artificial intelligence. It can be used across a range of tasks but has a particular focus on training and inference of deep neural networks. Researchers can develop machine learning and artificial intelligence-related applications with ease using comprehensive tools, libraries, and a community.

In this article, we will explain how to install TensorFlow on Ubuntu 20.04.

Create virtual environment

1. Update the package.
   

   $ sudo apt update 

2. Install supporting packages.
   

   $ sudo apt install software-properties-common

3. Add deadnsakes ppa with the below command.
   

   $ sudo add-apt-repository ppa:deadsnakes/ppa

4. Refresh the package list again.
   

   $ sudo apt update
   

5. Install python. Here we are Installing python 3.8.
   

   $ sudo apt install python3.8

6. Verify the python3 installation.
   

   $ python3 --version

   Output

   Python 3.8.10

7. Install Python virtual environment via the venv module on Ubuntu.
   

   $ sudo apt install python3-venv python3-dev

8. In your home directory, create a Tensor environment directory.
   

   $ mkdir tensor_env_directory

9. You can create a virtual environment with any name of your choice, we are creating a venv name  tensor_venv.

   $ cd tensor_env_directory

   $ python3 -m venv tensor_venv

10. After creating the environment, you will need to activate it using the below command.
    

    $ source tensor_venv/bin/activate

Upgrade pip

1. You will need to upgrade your pip installer before installing TensorFlow.
   

   $ pip install --upgrade pip

Install TensorFlow

1. The TensorFlow-CPU software package is simple to set up and supports CPU workloads.
   

   $ pip install tensorflow-cpu

2. Verify the TensorFlow installation.
   

   $ python3 -m pip show tensorflow-cpu

   Output

   
3. To verify the TensorFlow installation is working, Run the following command which generates random values.
   

   $ python -c "import tensorflow as tf; print(tf.random.normal([10,10]))"

   Output