Accelerated by another 5% using FP16/32 Batch-norm for Tensor Cores.

AlexeyAB

2018-04-16 9bae70b22549b68f5cdeece8b6c3b3de00c22714

 src/deconvolutional_layer.h

@@ -2,7 +2,6 @@
#define DECONVOLUTIONAL_LAYER_H

#include "cuda.h"
#include "params.h"
#include "image.h"
#include "activations.h"
#include "layer.h"