frederikfaye

Hi Taco, thanks for your answer! I feel like I didn't explain my point well enough the first time, so I will try to be explicit about the math: For...

Ah, that makes much more sense, thank you for the detailed explanation! (**EDIT**: I assume you meant "9\*4 parameters" instead of "9\*9\*4 parameters", since the example was using 3x3 filters.)...

Hi @marysia, thanks so much! Your second link is dead, I believe it should be https://github.com/marysia/GrouPy/tree/3d-gconv-python3 instead.

That `pip install splinterpy` is possible for Linux x86 users should probably be mentioned somewhere in splinter/docs/compile.md, as it is quite the time-saver.

@itsnamgyu Not just transfer learning, also when your own model consists of multiple submodels. Hope this will get fixed soon!

@martin-etchart Great that you got it to work! Where do you normally sequentially assemble your model? Could you share an example with what you got working? Thanks!

I would also love to have this initializer work out of the box. @iperov is suggesting changing https://github.com/keras-team/keras-contrib/blob/5ffab172661411218e517a50170bb97760ea567b/keras_contrib/initializers/convaware.py#L45 to `transpose_dimensions = (2, 3, 1, 0)`. Can anyone comment on what...

@jorgecarleitao Thank you for your answer! What if we consider the model I'm using a black box, that actually only takes the scalar variables as input, and which simply has...