Cyclops Tensor Framework
parallel arithmetic on multidimensional arrays
Model_trainer

Executes a set of different contractions on different processor counts to train model parameters. More...

Collaboration diagram for Model_trainer:

Namespaces

 CTF_int
 

Macros

#define TEST_SUITE
 

Functions

void train_off_vec_mat (int64_t n, int64_t m, World &dw, bool sp_A, bool sp_B, bool sp_C)
 
void train_ttm (int64_t sz, int64_t r, World &dw)
 
void train_dns_vec_mat (int64_t n, int64_t m, World &dw)
 
void train_sps_vec_mat (int64_t n, int64_t m, World &dw, bool sp_A, bool sp_B, bool sp_C)
 
void train_ccsd (int64_t n, int64_t m, World &dw)
 
void train_sparse_mp3 (int64_t n, int64_t m, World &dw)
 
void train_world (double dtime, World &dw, double step_size)
 
void frize (std::set< int > &ps, int p)
 
void train_all (double time, bool write_coeff, bool dump_data, std::string coeff_file, std::string data_dir)
 
char * getCmdOption (char **begin, char **end, const std::string &option)
 
int main (int argc, char **argv)
 

Detailed Description

Executes a set of different contractions on different processor counts to train model parameters.

Macro Definition Documentation

#define TEST_SUITE

Definition at line 10 of file model_trainer.cxx.

Function Documentation

void frize ( std::set< int > &  ps,
int  p 
)

Definition at line 280 of file model_trainer.cxx.

char* getCmdOption ( char **  begin,
char **  end,
const std::string &  option 
)

Definition at line 377 of file model_trainer.cxx.

Referenced by main().

int main ( int  argc,
char **  argv 
)
void train_all ( double  time,
bool  write_coeff,
bool  dump_data,
std::string  coeff_file,
std::string  data_dir 
)
void train_ccsd ( int64_t  n,
int64_t  m,
World dw 
)
void train_dns_vec_mat ( int64_t  n,
int64_t  m,
World dw 
)
void train_off_vec_mat ( int64_t  n,
int64_t  m,
World dw,
bool  sp_A,
bool  sp_B,
bool  sp_C 
)
void train_sparse_mp3 ( int64_t  n,
int64_t  m,
World dw 
)

Definition at line 216 of file model_trainer.cxx.

References sparse_mp3(), CTF::Timer::start(), and CTF::Timer::stop().

Referenced by train_world().

void train_sps_vec_mat ( int64_t  n,
int64_t  m,
World dw,
bool  sp_A,
bool  sp_B,
bool  sp_C 
)
void train_ttm ( int64_t  sz,
int64_t  r,
World dw 
)
void train_world ( double  dtime,
World dw,
double  step_size 
)