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

one more fix

Browse Source
This commit is contained in:
AlexeyAB
2018-02-23 00:13:08 +03:00
parent b2b5756d86
commit 6332ea99ab
1 changed files with 3 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/convolutional_kernels.cu
View File

@ -77,7 +77,8 @@ void binarize_weights_gpu(float *weights, int n, int size, float *binary)
__global__ void cuda_f32_to_f16(float* input_f32, size_t size, half *output_f16)
{
int idx = blockIdx.x * blockDim.x + threadIdx.x;
if (idx < size) output_f16[idx] = __float2half_rn(input_f32[idx]);
if (idx < size) output_f16[idx] = __float2half(input_f32[idx]);
//if (idx < size) *((unsigned int *)output_f16 + idx) = __float2half(input_f32[idx]);
}
void cuda_convert_f32_to_f16(float* input_f32, size_t size, half *output_f16) {
@ -88,6 +89,7 @@ __global__ void cuda_f16_to_f32(half* input_f16, size_t size, float *output_f32)
{
int idx = blockIdx.x * blockDim.x + threadIdx.x;
if (idx < size) output_f32[idx] = __half2float(input_f16[idx]);
//if (idx < size) output_f32[idx] = __half2float(*((unsigned int *)input_f16 + idx));
}
void cuda_convert_f16_to_f32(half* input_f16, size_t size, float *output_f32) {