Inserting batch normalisation layers

In Keras is there an easy way to insert layers at a given point in the model? According to the comments in the module, the .add() method is only good for putting layers on the end of the model.

I’m interested in adding batch normalisation to try and speed up the initial learning of my modified VGG model for the statefarm comp. If there’s no easy way then I can always pop() layers and add() them back again.

There’s not - but it’s something that I’m working on adding :slight_smile:

1 Like