Transfer Encoding Weights from Autoencoder to CNN

[maxkohlbrenner]

Implement a function in the convolutional autoencoder that stores the learned encoding weights and a function in the simple CNN that loads these encoding weights and uses them to initialize its own convolutional layers

[maxkohlbrenner]

Added to function stubs in simple_cnn.py and convolutional_autoencoder.py (store_encoding_weights, load_encoding_weights, marked with TODO)

Following steps are:

• implement both methods
• add a simple test (train a CAE very briefly similar to train_and_test_convolutional_autoencoder.py, save its encoding weights and load them into a cnn, visualize the wheights before storing them and after loading them and test whether everything went well)