GitHub-Proxy/eigen
1
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2025-06-04 18:54:00 +08:00
Code Issues Packages Projects Releases Wiki Activity

Improved the coverage of the fp16 reduction tests

Browse Source
This commit is contained in:
Benoit Steiner 2016-05-26 11:12:16 -07:00
parent 8288b0aec2
commit 22d02c9855
1 changed files with 26 additions and 42 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

68
unsupported/test/cxx11_tensor_of_float16_cuda.cu
View File

@ -248,65 +248,40 @@ void test_cuda_contractions() {
} }
void test_cuda_reductions() { void test_cuda_reductions(int size1, int size2, int redux) {
Eigen::CudaStreamDevice stream; Eigen::CudaStreamDevice stream;
Eigen::GpuDevice gpu_device(&stream); Eigen::GpuDevice gpu_device(&stream);
int size = 13; int num_elem = size1*size2;
int num_elem = size*size; int result_size = (redux == 1 ? size1 : size2);
float* d_float1 = (float*)gpu_device.allocate(num_elem * sizeof(float)); float* d_float1 = (float*)gpu_device.allocate(num_elem * sizeof(float));
float* d_float2 = (float*)gpu_device.allocate(num_elem * sizeof(float)); float* d_float2 = (float*)gpu_device.allocate(num_elem * sizeof(float));
Eigen::half* d_res_half = (Eigen::half*)gpu_device.allocate(size * sizeof(Eigen::half)); Eigen::half* d_res_half = (Eigen::half*)gpu_device.allocate(result_size * sizeof(Eigen::half));
Eigen::half* d_res_float = (Eigen::half*)gpu_device.allocate(size * sizeof(Eigen::half)); Eigen::half* d_res_float = (Eigen::half*)gpu_device.allocate(result_size * sizeof(Eigen::half));
Eigen::TensorMap<Eigen::Tensor<float, 2>, Eigen::Aligned> gpu_float1( Eigen::TensorMap<Eigen::Tensor<float, 2>, Eigen::Aligned> gpu_float1(
d_float1, size, size); d_float1, size1, size2);
Eigen::TensorMap<Eigen::Tensor<float, 2>, Eigen::Aligned> gpu_float2( Eigen::TensorMap<Eigen::Tensor<float, 2>, Eigen::Aligned> gpu_float2(
d_float2, size, size); d_float2, size1, size2);
Eigen::TensorMap<Eigen::Tensor<Eigen::half, 1>, Eigen::Aligned> gpu_res_half( Eigen::TensorMap<Eigen::Tensor<Eigen::half, 1>, Eigen::Aligned> gpu_res_half(
d_res_half, size); d_res_half, result_size);
Eigen::TensorMap<Eigen::Tensor<Eigen::half, 1>, Eigen::Aligned> gpu_res_float( Eigen::TensorMap<Eigen::Tensor<Eigen::half, 1>, Eigen::Aligned> gpu_res_float(
d_res_float, size); d_res_float, result_size);
gpu_float1.device(gpu_device) = gpu_float1.random(); gpu_float1.device(gpu_device) = gpu_float1.random();
gpu_float2.device(gpu_device) = gpu_float2.random(); gpu_float2.device(gpu_device) = gpu_float2.random();
Eigen::array<int, 1> redux_dim = {{0}}; Eigen::array<int, 1> redux_dim = {{redux}};
gpu_res_float.device(gpu_device) = gpu_float1.sum(redux_dim).cast<Eigen::half>(); gpu_res_float.device(gpu_device) = gpu_float1.sum(redux_dim).cast<Eigen::half>();
gpu_res_half.device(gpu_device) = gpu_float1.cast<Eigen::half>().sum(redux_dim); gpu_res_half.device(gpu_device) = gpu_float1.cast<Eigen::half>().sum(redux_dim);
Tensor<Eigen::half, 1> half_prec(size); Tensor<Eigen::half, 1> half_prec(result_size);
Tensor<Eigen::half, 1> full_prec(size); Tensor<Eigen::half, 1> full_prec(result_size);
gpu_device.memcpyDeviceToHost(half_prec.data(), d_res_half, size*sizeof(Eigen::half)); gpu_device.memcpyDeviceToHost(half_prec.data(), d_res_half, result_size*sizeof(Eigen::half));
gpu_device.memcpyDeviceToHost(full_prec.data(), d_res_float, size*sizeof(Eigen::half)); gpu_device.memcpyDeviceToHost(full_prec.data(), d_res_float, result_size*sizeof(Eigen::half));
gpu_device.synchronize(); gpu_device.synchronize();
for (int i = 0; i < size; ++i) { for (int i = 0; i < result_size; ++i) {
std::cout << "Checking redux " << i << std::endl;
VERIFY_IS_APPROX(full_prec(i), half_prec(i));
}
redux_dim = {{1}};
gpu_res_float.device(gpu_device) = gpu_float1.sum(redux_dim).cast<Eigen::half>();
gpu_res_half.device(gpu_device) = gpu_float1.cast<Eigen::half>().sum(redux_dim);
gpu_device.memcpyDeviceToHost(half_prec.data(), d_res_half, size*sizeof(Eigen::half));
gpu_device.memcpyDeviceToHost(full_prec.data(), d_res_float, size*sizeof(Eigen::half));
gpu_device.synchronize();
for (int i = 0; i < size; ++i) {
std::cout << "Checking redux " << i << std::endl;
VERIFY_IS_APPROX(full_prec(i), half_prec(i));
}
gpu_res_float.device(gpu_device) = gpu_float1.maximum(redux_dim).cast<Eigen::half>();
gpu_res_half.device(gpu_device) = gpu_float1.cast<Eigen::half>().maximum(redux_dim);
gpu_device.memcpyDeviceToHost(half_prec.data(), d_res_half, size*sizeof(Eigen::half));
gpu_device.memcpyDeviceToHost(full_prec.data(), d_res_float, size*sizeof(Eigen::half));
gpu_device.synchronize();
for (int i = 0; i < size; ++i) {
std::cout << "Checking redux " << i << std::endl; std::cout << "Checking redux " << i << std::endl;
VERIFY_IS_APPROX(full_prec(i), half_prec(i)); VERIFY_IS_APPROX(full_prec(i), half_prec(i));
} }
@ -317,7 +292,16 @@ void test_cuda_reductions() {
gpu_device.deallocate(d_res_float); gpu_device.deallocate(d_res_float);
} }
void test_cuda_reductions() {
test_cuda_reductions(13, 13, 0);
test_cuda_reductions(13, 13, 1);
test_cuda_reductions(35, 36, 0);
test_cuda_reductions(35, 36, 1);
test_cuda_reductions(36, 35, 0);
test_cuda_reductions(36, 35, 1);
}
void test_cuda_full_reductions() { void test_cuda_full_reductions() {
Eigen::CudaStreamDevice stream; Eigen::CudaStreamDevice stream;
@ -427,8 +411,8 @@ void test_cxx11_tensor_of_float16_cuda()
CALL_SUBTEST_1(test_cuda_trancendental()); CALL_SUBTEST_1(test_cuda_trancendental());
CALL_SUBTEST_2(test_cuda_contractions()); CALL_SUBTEST_2(test_cuda_contractions());
CALL_SUBTEST_3(test_cuda_reductions()); CALL_SUBTEST_3(test_cuda_reductions());
CALL_SUBTEST_3(test_cuda_full_reductions()); CALL_SUBTEST_4(test_cuda_full_reductions());
CALL_SUBTEST_4(test_cuda_forced_evals()); CALL_SUBTEST_5(test_cuda_forced_evals());
#else #else
std::cout << "Half floats are not supported by this version of cuda: skipping the test" << std::endl; std::cout << "Half floats are not supported by this version of cuda: skipping the test" << std::endl;
#endif #endif