AIfES 2
2.0.0
|
General Stochastic Gradient Descent (SGD) optimizer struct. More...
#include <aiopti_sgd.h>
Data Fields | |
aiopti_t | base |
Inherited field members from general optimizer struct. | |
Optimizer configuration | |
Required configuration parameters for the optimizer These fields have to be configured by the user before calling the initializer function. | |
void * | momentum |
aiscalar: Momentum(set to null to save optimization memory) | |
Math functions | |
Required data type specific math functions | |
void(* | tensor_sub )(const aitensor_t *a, const aitensor_t *b, aitensor_t *result) |
Required math function: Element wise tensor subtraction. More... | |
void(* | tensor_add )(const aitensor_t *a, const aitensor_t *b, aitensor_t *result) |
Required math function: Element wise tensor addition. More... | |
void(* | scalar_mul )(const void *scalar, const aitensor_t *a, aitensor_t *result) |
Required math function: Multiplication of a scalar with a tensor. More... | |
void(* | zero_tensor )(aitensor_t *tensor) |
Required math function: Sets the elements of a tensor to zero. More... | |
General Stochastic Gradient Descent (SGD) optimizer struct.
void(* scalar_mul) (const void *scalar, const aitensor_t *a, aitensor_t *result) |
Required math function: Multiplication of a scalar with a tensor.
Requires a math function that multiplies a scalar with a tensor:
\[ result = scalar \cdot a \]
void(* tensor_add) (const aitensor_t *a, const aitensor_t *b, aitensor_t *result) |
Required math function: Element wise tensor addition.
Requires a math function that adds two tensors element wise:
\[ result = a + b \]
void(* tensor_sub) (const aitensor_t *a, const aitensor_t *b, aitensor_t *result) |
Required math function: Element wise tensor subtraction.
Requires a math function that subtracts two tensors element wise:
\[ result = a - b \]
void(* zero_tensor) (aitensor_t *tensor) |
Required math function: Sets the elements of a tensor to zero.
Requires a math function that sets the elements of the given tensor to zero (or to the zero_point):
\[ tensor_i = 0 \]