Installing Keras Using Python & TensorFlow: A Guide
What is Keras?
Keras is a tool for machine learning specialists who work with Python, mostly used due to the convenience of mathematical calculations. Developers use Keras to create, configure, and test machine learning and artificial intelligence systems, primarily neural networks.
What are Keras Models?
Keras works with models or schemes by which information is distributed and transformed. Machine learning is essentially processing information using a programmed network, where certain conclusions are drawn based on certain data. The network structure is called a model and is often presented as a graph, diagram, or table.
Requirements
- A server with root-level access.
- This tutorial is performed with CentOS 7.
- Python installed.
- TensorFlow installed.
Installing Python on CentOS
Step 1:
Log in to your CentOS system as a root user or a user with sudo privileges.
# sudo su -
You can also log in via secure shell (SSH) using the following command. Replace username with your username and serveripaddress with your server’s IP address.
# ssh username@serveripaddress
Step 2:
Update the system packages to avoid errors. This tutorial uses the Yum package manager.
# yum update -y
Step 3:
CentOS SCL is a tool that allows you to install updated versions of the same software. Install CentOS SCL using this command.
# sudo yum install centos-release-scl
Python is typically installed by default on most systems. To check, verify the version using the following command.
# python --version
You probably see a Python version 2.X.X. Keep this installed on your machine as other software may use this particular Python version.
Step 4:
Tensorflow and Keras require a python version of at least 3.7. Use the following command to install Python 3.8.
# sudo yum install rh-python38
After installing Python 3.8, assign this version as the standard version.
# scl enable rh-python38 bash
Step5:
Recheck the version to ensure 3.8 is the standard version.
# python --version
Installing TensorFlow
Step1:
You need to create a virtual environment when installing TensorFlow. First, create a folder for your project using the following commands.
# mkdir test
# cd test
Use the following command to create a virtual environment with python. It creates a virtual environment named tf-virtual-env. Replace this with your chosen name.
# python3 -m venv tf-virtual-env
Inside this environment are many pre-installed Python libraries and tools needed in the project, such as Package installer for Python (pip). The following command activates the environment.
# source tf-virtual-env/bin/activate
The prompt changes in the terminal, which means you activated it successfully.
(tf-virtual-env) # _
Type deactivate at the prompt and press Enter to exit the environment.
TensorFlow requires at least version 19.0 of pip. Use the following command to ensure you have the latest version.
# pip install --upgrade pip
Step 2:
This tutorial installs a version that does not use your GPU. You can install a TensorFlow version that offers GPU support. Use the following command to install TensorFlow without GPU support.
# pip install --upgrade tensorflow
Use the same command for updating TensorFlow.
Step 3:
To test your environment, open Python bash.
# python
Then, type the following command. The first line will import the TensorFlow packages into the Python interpreter session, while the second line will print the TensorFlow version.
import tensorflow as tf
print(tf.__version__)'
Install Keras
Thanks to a new update in TensorFlow 2.0+, if you installed TensorFlow as instructed, you don’t need to install Keras anymore because it is installed with TensorFlow.
To confirm it, open Python bash.
# python
At the prompt, run the following commands.
import keras
keras.__version__
For those using TensorFlow versions before 2.0, here are the instructions for installing Keras using pip.
Pip Install
Before beginning, update pip.
# pip install --upgrade pip
Then, install Keras with the following command.
# pip install keras
Deprecation of the Git Clone Keras Install Method
Keras was previously installed by cloning the GitHub repository, unpacking the packages, and installing the software. The Keras team deprecated the GitHub repository and moved the applications into the core Keras repository and the TensorFlow pip package.

The recommended Keras installation method from the Keras team is via TensorFlow version 2+.
Importing Keras Libraries
When we create a python virtual environment, it already contains most of the important libraries. Here is the list of the most important Python libraries:
- Numpy.
- Pandas.
- Scikit-learn.
- Matplotlib.
- Scipy.
- Seabord.
To confirm the Numpy library, use the following command.
import numpy as np
ndarray = np.ones([3, 3])
print("TensorFlow operations convert numpy arrays to Tensors automatically")
tensor = tf.math.multiply(ndarray, 42)
print(tensor)
The output should be as follows.
tf.Tensor( [[42. 42. 42.]
[42. 42. 42.]
[42. 42. 42.]], shape=(3, 3), dtype=float64)
Updating Keras
Since Keras installs alongside Tensorflow, they also update together. Use the command below, the same one to update TensorFlow, to update Keras.
# pip install --upgrade tensorflow
Wrapping Up
There are ways to install Karas and Tensorflow without a virtual environment. Still, it can be risky and more difficult than the commands of pre-configured environments. Whether installing Keras using pip via Python or TensorFlow, this tutorial helps you get it up and running for your next deep learning project.
Do you need a CentOS or AlmaLinux machine for your next Keras install? Liquid Web these options for VPS Hosting, Cloud Dedicated Servers, and Dedicated Servers. Contact our sales team today for setup options.
Related Articles:
- How to Edit the PHP Memory for Your WordPress Site via WP Toolkit
- 4 Methods for How to Install Yarn on Windows Server
- How to Install Bpytop Resource Monitoring Tool on AlmaLinux
- How to Fix “This Site Can’t Provide a Secure Connection” Error
- How to Install MongoDB on AlmaLinux
- How to Install PostgreSQL on AlmaLinux

About the Author: Zachary Armstrong
Zachary is a writer who specializes in breaking down complex subjects and making them easy to understand. He has a passion for technology, believes it can change the world for the better, and wants to tell the whole world about it.
Our Sales and Support teams are available 24 hours by phone or e-mail to assist.
Latest Articles
How to Edit the PHP Memory for Your WordPress Site via WP Toolkit
Read ArticleWhat is CGI-Bin and What Does it Do?
Read ArticleTop 10 Password Security Standards
Read ArticleTop 10 Password Security Standards
Read ArticleHow to Use the WP Toolkit to Secure and Update WordPress
Read Article