GitLab

in-place.

that contain a different number of samples than their input tensors.

vector from that layer so that it can be reused in any computations.
Most layers won't use it but for some backward computations having it
allows them to be implemented more efficiently.

to just regular lists of parameters.

cuDNN.

more of the layers_abstract.h file.

have a tag that refers to the raw input given to a network.