I'm using a Convolutional Network, and I wonder which data_format
is being used in my code? Because from this page: Convolutional Layers
In the Conv2D layer section, it says input shape can assume two different forms depending on the data_format
("channels_first"
or "channels_last"
). I assume on of them is the default option?
And what command should I add to my code to change between those two formats?
Oh, just found the answer, here:
So, it uses the "channels_last"
if using TensorFlow backend to keras and "channels_first"
if using Theano backend.