cu-rand-speed-test.cc File Reference

#include <iostream>
#include <vector>
#include <cstdlib>
#include "base/kaldi-common.h"
#include "util/common-utils.h"
#include "cudamatrix/cu-matrix.h"
#include "cudamatrix/cu-vector.h"
#include "cudamatrix/cu-rand.h"

Include dependency graph for cu-rand-speed-test.cc:

Go to the source code of this file.

Namespaces
	kaldi
	This code computes Goodness of Pronunciation (GOP) and extracts phone-level pronunciation feature for mispronunciations detection tasks, the reference:

Functions
template<typename Real >
std::string	NameOf ()
template<typename T >
std::string	ToString (const T &t)
template<typename Real >
std::string	MeanVariance (const CuMatrixBase< Real > &m)
template<typename Real >
std::string	MeanVariance (const CuVectorBase< Real > &v)
template<typename Real >
void	CuRandUniformMatrixSpeedTest (const int32 iter)
template<typename Real >
void	CuRandUniformMatrixBaseSpeedTest (const int32 iter)
template<typename Real >
void	CuRandGaussianMatrixSpeedTest (const int32 iter)
template<typename Real >
void	CuRandGaussianMatrixBaseSpeedTest (const int32 iter)
template<typename Real >
void	CuRandUniformVectorSpeedTest (const int32 iter)
template<typename Real >
void	CuRandGaussianVectorSpeedTest (const int32 iter)
int	main ()

Function Documentation

main()

int main

(

)

Definition at line 188 of file cu-rand-speed-test.cc.

References Timer::Elapsed(), and KALDI_LOG.

           {
  int32 iter = 10; // Be quick on CPU,
#if HAVE_CUDA == 1
  for (int32 loop = 0; loop < 2; loop++) { // NO for loop if 'HAVE_CUDA != 1',
    CuDevice::Instantiate().SetDebugStrideMode(true);
    if ( loop == 0)
      CuDevice::Instantiate().SelectGpuId("no");
    else {
      CuDevice::Instantiate().SelectGpuId("yes");
      iter = 400; // GPUs are faster,
    }
#endif
    Timer t;
    kaldi::CuRandUniformMatrixSpeedTest<float>(iter);
    kaldi::CuRandUniformMatrixBaseSpeedTest<float>(iter);
    kaldi::CuRandUniformVectorSpeedTest<float>(iter);
    kaldi::CuRandGaussianMatrixSpeedTest<float>(iter);
    kaldi::CuRandGaussianMatrixBaseSpeedTest<float>(iter);
    kaldi::CuRandGaussianVectorSpeedTest<float>(iter);
    fprintf(stderr, "---\n");

    kaldi::CuRandUniformMatrixSpeedTest<double>(iter);
    kaldi::CuRandUniformMatrixBaseSpeedTest<double>(iter);
    kaldi::CuRandUniformVectorSpeedTest<double>(iter);
    kaldi::CuRandGaussianMatrixSpeedTest<double>(iter);
    kaldi::CuRandGaussianMatrixBaseSpeedTest<double>(iter);
    kaldi::CuRandGaussianVectorSpeedTest<double>(iter);
    fprintf(stderr, "--- ELAPSED %fs.\n\n", t.Elapsed());
#if HAVE_CUDA == 1
  } // No for loop if 'HAVE_CUDA != 1',
  CuDevice::Instantiate().PrintProfile();
#endif
  KALDI_LOG << "Tests succeeded.";
}