Follow the installation instructions on TensorFlow.org to create a Virtual Environment for your platform, but stop before installing TensorFlow with pip (e.g. do not run this command pip install --upgrade tensorflow )
After activating your virtual environment, run these commands inside it:
$ git clone https://github.com/random-forests/tensorflow-workshop.git
$ cd tensorflow-workshop
$ pip install -r setup/requirements.txt
This will install the correct TensorFlow version and other libraries within your virtual environment. Now, start a Jupyter Notebook server by running:
$ jupyter notebook
If you experience difficulty, try the Docker option on the previous page for an easier option, or see the full installation instructions on tensorflow.org.