A large number of docker images and machine images with preconfigured libraries and APIs are available at AWS marketplace. To launch the jupyter notebook , at command line use <jupyter notebook --ip=0.0.0.0 --no-browser>. This will result in an output as shown in the following:

Copy/paste this URL into your browser when you connect for the first time to login with a token:
http://0.0.0.0:8888/?token=3156e...

You copy and paste the URL on your browser and you are ready to go.

In addition, the whole process can be simplified by looking at AWS CloudFormation. It creates and configures Amazon Web Services resources with a template. In doing so, it simplifies the process of setting up a distributed deep learning cluster. The interested reader can have a look to https://aws.amazon.com/blogs/compute/distributed-deep-learning-made-easy/