Tensor Flow Extended Tutorials

First, we need to ensure you have a recent version of Anaconda installed:

• If you don't have Anaconda, follow this Anaconda installation tutorial.

• If you already have Anaconda: On Windows open the Start menu and open an Anaconda Command Prompt. On macOS or Linux open a terminal window:

conda update --all

• In your terminal: cd into a folder you want to install all demos.

cd your-path

• Clone the repository containing all demos

git clone https://github.com/kirenz/tfx

• cd into the tfx folder

cd tfx

• Now we install all needed Python libraries from the environment.yml file to create a virtual environment named tfx-env. This includes the latest version of Tensor Flow and Tensor Flow Extended:

conda env create -f environment.yml

• Activate the tfx-env environment:

conda activate tfx-env

• Check your version of TFX with tfx_check.ipynb.

• Go to the folder penguin_template and use the starter code in penguin_simple.ipynb to build your TFX-pipeline.

• The folder penguin_solution includes a succesfull run of the penguin_simple.ipynb file.

Code

1. Penguin_simple
2. Penguin_tfdv