Make sure you have python3 installed.

python --version

If this returns Python 2.X.X then try this:

python3 --version

This should return Python 3.X.X. If it fails then you will need to install Python3. Follow the instructions for your operating system to install it.

We will be using a virtual environment for these tutorials. This keeps all the dependencies nicely isolated from the rest of your python installation.

This will create a virtual environment:

python3 -m venv venv

We then activate our virtual environment using:

. ./venv/bin/activate

It's also worth making sure pip is up to date:

pip install --upgrade pip

And then we install our dependencies using:

pip install -r requirements.txt

Activate the virtual environment that you created in the Setup steps:

. ./venv/bin/activate

And then run:

jupyter notebook .

You can now open the various notebooks.

This notebook demonstrates how to classify inputs into 2 categories, True or False.

This notebook demonstrates how to classify inputs into multiple potential categories - e.g. Cat, Dog, Horse, Cow...

This notebook demonstrates how to use the Sparse Categorical Crossentropy loss function which is similar to the Categorical Crossentropy cost function but removes the need for one-hot encoding the training labels.