Mathis/CUDANet
1
0
Fork 0
mirror of https://github.com/lordmathis/CUDANet.git synced 2025-11-06 01:34:22 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix conv2d kernel dims

Browse Source
This commit is contained in:
Matúš Námešný
2024-03-09 22:55:37 +01:00
parent 96804777ee
commit 4f3c4f1afb
1 changed files with 4 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/layers/conv2d.cu
View File

@@ -32,13 +32,13 @@ Layers::Conv2d::Conv2d(
outputSize = (inputSize - kernelSize) / stride + 1;
}
kernels.resize(kernelSize * kernelSize * numFilters);
kernels.resize(kernelSize * kernelSize * inputChannels * numFilters);
initializeKernels();
d_kernels = nullptr;
CUDA_CHECK(
cudaMalloc((void**)&d_kernels, sizeof(float) * kernelSize * kernelSize * numFilters)
cudaMalloc((void**)&d_kernels, sizeof(float) * kernelSize * kernelSize * inputChannels * numFilters)
);
toCuda();
@@ -84,6 +84,8 @@ void Layers::Conv2d::forward(const float* d_input, float* d_output) {
convolution_kernel<<<1, THREADS_PER_BLOCK>>>(
d_padded, d_kernels, d_output, inputSize + (2 * paddingSize), inputChannels, kernelSize, stride, numFilters, outputSize
);
CUDA_CHECK(cudaDeviceSynchronize());
}
/*