To help you understand the concepts easily, most of the examples have two parts, the first being the native Tensorflow version on centerized plaintext datasets, the second being the Rosetta version, which has the same functionality while running on secret-sharing 'ciphertext' datasets in a privacy-preserving way.
In order to run the examples smoothly, some dependencies need to be installed, as follows:
pip3 install matplotlib sklearn --userEnjoy it.