This will guide you through how to pull the docker container for TensorFlow with Python 3 and Jupyter Notebooks on Windows.
We assume you have docker desktop installed.
Docker simplifies the process of getting software running on your machine by providing a virtual computing environment. However, we want to save our work locally which requires mounting a local directory to our container.
First, create a directory in which to save your notebooks.
mkdir my_jupyter_notebooks
Next we need to install and run our container. This may take a little while the first time.
docker run -it -p 8889:8888 -v ./my_jupyter_notebooks:/tf tensorflow/tensorflow:latest-py3-jupyter
When the container starts, you should see a screen with with a bunch of information about your Jupyter server. All of the urls that you see should have a token that you will need. Copy it to your clipboard.
In a webbrowser, navigate to 'localhost:8889' and enter your token in the password.
This brings you to the tensorflow tutorials section.
Congratulations! You've successfully started your Docker Container.
Other commands...
docker inspect -f '{{range.NetworkSettings.Networks}}{{.IPAddress}}{{end}}' container_name_or_id