From 9bae70b22549b68f5cdeece8b6c3b3de00c22714 Mon Sep 17 00:00:00 2001
From: AlexeyAB <alexeyab84@gmail.com>
Date: Mon, 16 Apr 2018 23:51:11 +0000
Subject: [PATCH] Accelerated by another 5% using FP16/32 Batch-norm for Tensor Cores.
---
src/normalization_layer.c | 6 ++++++
1 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/src/normalization_layer.c b/src/normalization_layer.c
index 0551337..069a079 100644
--- a/src/normalization_layer.c
+++ b/src/normalization_layer.c
@@ -21,7 +21,13 @@
layer.norms = calloc(h * w * c * batch, sizeof(float));
layer.inputs = w*h*c;
layer.outputs = layer.inputs;
+
+ layer.forward = forward_normalization_layer;
+ layer.backward = backward_normalization_layer;
#ifdef GPU
+ layer.forward_gpu = forward_normalization_layer_gpu;
+ layer.backward_gpu = backward_normalization_layer_gpu;
+
layer.output_gpu = cuda_make_array(layer.output, h * w * c * batch);
layer.delta_gpu = cuda_make_array(layer.delta, h * w * c * batch);
layer.squared_gpu = cuda_make_array(layer.squared, h * w * c * batch);
--
Gitblit v1.10.0