Pass the component initializer as an hparam slows down the script very much. The
API has now been changed to pass it as a kwarg to the models instead.
The example scripts have also been updated to reflect the new changes.
Also, ImageGMLVQ and an example script `gmlvq_mnist.py` that uses it have also
been added.
How to reproduce:
Run the `glvq_spiral.py` file under `examples/`.
The error seems to occur when using a lot of prototypes in combination with the
`StratifiedSelectionInitializer`. Using only a prototype per class, or using
another initializer like the `StratifiedMeanInitializer` seems to make the
problem go away.