mirror/darknet
1
0
Fork 0
mirror of https://github.com/pjreddie/darknet.git synced 2023-08-10 21:13:14 +03:00
Code Issues Projects Releases Wiki Activity

Fix OpenCV compilation. maxpool_layer on GPU uses cuDNN.

Browse Source
This commit is contained in:
AlexeyAB
2018-10-10 13:51:36 +03:00
parent dbc75aec77
commit 7ee4135910
6 changed files with 55 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
src/maxpool_layer.c
View File

@ -47,7 +47,27 @@ maxpool_layer make_maxpool_layer(int batch, int h, int w, int c, int size, int s
l.indexes_gpu = cuda_make_int_array(output_size);
l.output_gpu = cuda_make_array(l.output, output_size);
l.delta_gpu = cuda_make_array(l.delta, output_size);
#endif
#ifdef CUDNN
cudnnStatus_t maxpool_status;
maxpool_status = cudnnCreatePoolingDescriptor(&l.poolingDesc);
maxpool_status = cudnnSetPooling2dDescriptor(
l.poolingDesc,
CUDNN_POOLING_MAX,
CUDNN_PROPAGATE_NAN, // CUDNN_PROPAGATE_NAN, CUDNN_NOT_PROPAGATE_NAN
l.size,
l.size,
0, //l.pad,
0, //l.pad,
l.stride,
l.stride);
cudnnCreateTensorDescriptor(&l.srcTensorDesc);
cudnnCreateTensorDescriptor(&l.dstTensorDesc);
cudnnSetTensor4dDescriptor(l.srcTensorDesc, CUDNN_TENSOR_NCHW, CUDNN_DATA_FLOAT, l.batch, l.c, l.h, l.w);
cudnnSetTensor4dDescriptor(l.dstTensorDesc, CUDNN_TENSOR_NCHW, CUDNN_DATA_FLOAT, l.batch, l.out_c, l.out_h, l.out_w);
#endif // CUDNN
#endif // GPU
l.bflops = (l.size*l.size*l.c * l.out_h*l.out_w) / 1000000000.;
fprintf(stderr, "max %d x %d / %d %4d x%4d x%4d -> %4d x%4d x%4d %5.3f BF\n", size, size, stride, w, h, c, l.out_w, l.out_h, l.out_c, l.bflops);
return l;