Introduction to Deep Learning with Keras (Part 1): Installing Keras

Starting to learn deep learning can be a really intimidating task. Most people start with traditional machine learning with Scikit-Learn. Then they slowly start to move towards deep learning trying to implement whatever they have learned. Again, many of them start with TensorFlow directly. But it can be really difficult when using only TensorFlow in the learning phase. And for that reason specifically, we have Keras which makes it a lot easier to dive into Deep Learning.

This is the first part of the series Introduction to Keras Deep Learning.
Part 1: Getting Started with Keras.

In this article, we will be covering the basics of Keras API and its installation procedure.

What will You Learn in this Article?

• What is Keras?
• How to install Keras on your system?

What is Keras?

Keras is a high-level API for deep learning. It can run on top of TensorFlow, CNTK, and Theano as well. The main focus for developing a high-level API like Keras was fast prototyping and implementation. But we can do a lot more than just prototyping in Keras. Once you start to play around, you will get a sense of how powerful it is.

Keras has become so popular that TensorFlow has implemented it as an internal feature. You can implement Keras API through TensorFlow by using tf.keras module. But for this series, we will stick to the standalone Keras API.

How to Install Keras in Your System?

Keras is the high-level API and to use it you need to install at least one of the following backend engines:
1. TensorFlow
2. Theano
3. CNTK

Now, its time to install Keras. There are two convenient ways to do so.

1. Install Keras using pip Package Manager:

To install Keras using the pip package manager type the following:

pip install keras

2. Install Keras using the conda Command:

To use the Anaconda Prompt, first, you will need to install Anaconda. You can follow this article (Python Environment Setup for Deep Learning) to install Anaconda first.

This is a more convenient way because anaconda installs many of the optional packages as well including the compatible version of cuDNN. Installing Anaconda will also help you to create a virtual environment with much ease.

After you have installed Anaconda, type the following command:

conda install keras

That’s it. Now you are all set to learn Deep Learning with Keras. Obviously, there are many more packages that we will be needing through this tutorial series. Although Anaconda installs most of them, we will still go through the installation procedure of each of them when they come up during the series.

Summary and Conclusion

It is better to keep the setup tutorials and working code tutorials in separate articles. We will be covering many of Keras’ important and commonly used functions and imports in the next tutorial of the series. If you have any queries, ask them in the comment section.

If you liked this article then comment, share and give a thumbs up. If you have any questions or suggestions, just Contact me here. Be sure to subscribe to the website for more content. Follow me on Twitter, LinkedIn, and Facebook to get regular updates.