SPRAAK
 All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Groups Pages
Data Structures | Macros | Typedefs | Functions | Variables
API_LowLvl::lib::nn
+ Collaboration diagram for API_LowLvl::lib::nn:

Data Structures

struct  SprNNOptions
 

Macros

#define SPR_NN_VEC_SZ
 
#define SPR_NN_VEC_AS
 
#define SprNNFltV_
 
#define SprNNFltA_
 
#define SprNNIntV_
 
#define SprNNIntA_
 
#define SprNNUIntV_
 
#define SprNNUIntA_
 
#define SPR_NN_N_PARALLEL
 compute SPR_NN_N_PARALLEL input together in 'parallel' mode (faster) More...
 
#define spr_nn_dt_flt
 

Typedefs

typedef float SprNNFltS
 a single floating point parameter More...
 
typedef int32_t SprNNIntS
 this type encodes the corresponding int type More...
 
typedef uint32_t SprNNUIntS
 this type encodes the corresponding int type More...
 
typedef SprNNFltV_ SprNNFltV
 a vector of floating point parameters More...
 
typedef SprNNIntV_ SprNNIntV
 this type encodes the corresponding int type More...
 
typedef SprNNUIntV_ SprNNUIntV
 this type encodes the corresponding unsigned int type More...
 
typedef SprNNFltA_ SprNNFltA
 
typedef SprNNIntA_ SprNNIntA
 this type encodes the corresponding int type More...
 
typedef SprNNUIntA_ SprNNUIntA
 this type encodes the corresponding unsigned int type More...
 

Functions

unsigned int spr_nn_decode_options (const char *desc, int *MT)
 

Variables

const SprCmdOptDesc spr_nn_option_desc []
 

Detailed Description

Macro Definition Documentation

#define SPR_NN_VEC_SZ
#define SPR_NN_VEC_AS
#define SprNNFltV_
#define SprNNFltA_
#define SprNNIntV_
#define SprNNIntA_
#define SprNNUIntV_
#define SprNNUIntA_
#define SPR_NN_N_PARALLEL

compute SPR_NN_N_PARALLEL input together in 'parallel' mode (faster)

tested with 4 only!

#define spr_nn_dt_flt

Typedef Documentation

typedef float SprNNFltS

a single floating point parameter

the basic type used to evaluate all operations

typedef int32_t SprNNIntS

this type encodes the corresponding int type

typedef uint32_t SprNNUIntS

this type encodes the corresponding int type

a vector of floating point parameters

the type used to speed up to computations, encodes the minimum required data alignment

this type encodes the corresponding int type

this type encodes the corresponding unsigned int type

this type encodes the optimal data alignment when using a vector of the type SprNNFltV

this type encodes the corresponding int type

this type encodes the corresponding unsigned int type

Function Documentation

unsigned int spr_nn_decode_options ( const char *  desc,
int *  MT 
)

Decode the textual NN options given in desc and decode into flags (return value) and the number of threads *MT.

Returns
the decoded flags and (~0u) on error.

Variable Documentation

const SprCmdOptDesc spr_nn_option_desc[]