shortcut layers, msr networks

Joseph Redmon

2015-12-14 db0397cfaaf488364e3d2e1669dfefae2ee6ea73

 src/convolutional_kernels.cu

@@ -122,6 +122,12 @@
            fast_mean_gpu(l.output_gpu, l.batch, l.n, l.out_h*l.out_w, l.mean_gpu);
            fast_variance_gpu(l.output_gpu, l.mean_gpu, l.batch, l.n, l.out_h*l.out_w, l.variance_gpu);

            /*
            cuda_pull_array(l.variance_gpu, l.mean, 1);
            printf("%f\n", l.mean[0]);
            */


            scal_ongpu(l.n, .95, l.rolling_mean_gpu, 1);
            axpy_ongpu(l.n, .05, l.mean_gpu, 1, l.rolling_mean_gpu, 1);
            scal_ongpu(l.n, .95, l.rolling_variance_gpu, 1);