[ GPU ] split kernel registration from forwarding function in addition_layer_cl
and transpose_cl
#2810
Loading
addition_layer_cl
and transpose_cl
#2810