![]() |
MMALIB User Guide
|
Structure containing the parameters initialization of CNN convolution computation.
Definition at line 99 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
Data Fields | |
int8_t | funcStyle |
Variant of the function refer to MMALIB_FUNCTION_STYLE More... | |
int32_t | No |
Number of output feature maps More... | |
int32_t | inChOffset |
offset of the input feature maps in B matrix. This is power of 2 for circular buffer and atleast 64 byte aligned for linear buffer More... | |
int32_t | validColsIn |
Valid columns of input feature maps in B matrix for one call of processing of the kernel for non strided convolution but including 1x1 stride 2. This is in units of data type More... | |
int32_t | validColsPerRowIn |
Valid columns in a row of input feature maps for one call of processing of the kernel for non strided convolution and for 1x1 stride 2. This will not be used in J7AM instead use inWidth More... | |
int32_t | validRowsIn |
Valid input rows of input feature maps for one call of processing of the kernel for strided convolution except for 1x1 stride 2 More... | |
int32_t | inputPitchPerRow |
Valid pitch for each input rows of input feature maps for strided convolution. This is units of bytes More... | |
int32_t | outputPitchPerRow |
Valid output pitch for each output rows of output feature maps for strided convolution. This is units of bytes More... | |
int32_t | inWidth |
Width of each row of input feature map in units of data type More... | |
int32_t | pad |
Pad of each row of input feature map, specify pad only one one side. specific to J7ES More... | |
int32_t | padLeft |
Pad of each row of input feature map, specify pad only on left side More... | |
int32_t | padRight |
Pad of each row of input feature map, specify pad only on right side More... | |
int32_t | padTop |
Pad of each row of input feature map, specify pad only on top side More... | |
int32_t | padBottom |
Pad of each row of input feature map, specify pad only on bottom side More... | |
int32_t | validColsOutBottom |
Output pixels for stride 1 flow when padBotton > 0 More... | |
int32_t | maxHeight |
Height of the input feature map in units of data type More... | |
int32_t | subMChannels |
Number of output channels per kernel call More... | |
int32_t | numGroupsPerKernel |
Scale parameter for output precision / uint8_t scale; /! More... | |
int32_t | Fr |
coefficient rows (height) More... | |
int32_t | Fc |
coefficient columns (width) More... | |
int32_t | strideX |
stride of columns More... | |
int32_t | strideY |
stride of rows More... | |
int32_t | dilationX |
dilation of coefficients of columns More... | |
int32_t | dilationY |
dilation of coefficients of rows More... | |
int32_t | bias |
bias value in B matrix same as data type of B matrix More... | |
uint8_t | activationType |
activation RELU, SAT or none for output More... | |
int32_t | pSatMin |
Min value for parameterized saturation More... | |
int32_t | pSatMax |
Max value for parameterized saturation More... | |
uint8_t | mode |
mode for input feature map in Circular or Linear mode in B matrix More... | |
int32_t | col |
column number of input feature map More... | |
int8_t | packetizeMode |
packetizeMode is 0 then weights are not packetized. When 1 then they are packetized More... | |
int8_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::funcStyle |
Variant of the function refer to MMALIB_FUNCTION_STYLE
Definition at line 102 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::No |
Number of output feature maps
Definition at line 104 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::inChOffset |
offset of the input feature maps in B matrix. This is power of 2 for circular buffer and atleast 64 byte aligned for linear buffer
Definition at line 107 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::validColsIn |
Valid columns of input feature maps in B matrix for one call of processing of the kernel for non strided convolution but including 1x1 stride 2. This is in units of data type
Definition at line 111 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::validColsPerRowIn |
Valid columns in a row of input feature maps for one call of processing of the kernel for non strided convolution and for 1x1 stride 2. This will not be used in J7AM instead use inWidth
Definition at line 115 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::validRowsIn |
Valid input rows of input feature maps for one call of processing of the kernel for strided convolution except for 1x1 stride 2
Definition at line 119 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::inputPitchPerRow |
Valid pitch for each input rows of input feature maps for strided convolution. This is units of bytes
Definition at line 122 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::outputPitchPerRow |
Valid output pitch for each output rows of output feature maps for strided convolution. This is units of bytes
Definition at line 125 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::inWidth |
Width of each row of input feature map in units of data type
Definition at line 127 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::pad |
Pad of each row of input feature map, specify pad only one one side. specific to J7ES
Definition at line 130 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::padLeft |
Pad of each row of input feature map, specify pad only on left side
Definition at line 133 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::padRight |
Pad of each row of input feature map, specify pad only on right side
Definition at line 136 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::padTop |
Pad of each row of input feature map, specify pad only on top side
Definition at line 139 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::padBottom |
Pad of each row of input feature map, specify pad only on bottom side
Definition at line 142 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::validColsOutBottom |
Output pixels for stride 1 flow when padBotton > 0
Definition at line 144 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::maxHeight |
Height of the input feature map in units of data type
Definition at line 146 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::subMChannels |
Number of output channels per kernel call
Definition at line 148 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::numGroupsPerKernel |
Scale parameter for output precision / uint8_t scale; /!
Shift parameter for output precision / int32_t shift; /! number of groups per kernel call > 1 will enable processing when No <= MMA size for non strided kernels and default value is 1
Definition at line 155 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::Fr |
coefficient rows (height)
Definition at line 157 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::Fc |
coefficient columns (width)
Definition at line 159 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::strideX |
stride of columns
Definition at line 161 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::strideY |
stride of rows
Definition at line 163 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::dilationX |
dilation of coefficients of columns
Definition at line 165 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::dilationY |
dilation of coefficients of rows
Definition at line 167 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::bias |
bias value in B matrix same as data type of B matrix
Definition at line 169 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
uint8_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::activationType |
activation RELU, SAT or none for output
Definition at line 171 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::pSatMin |
Min value for parameterized saturation
Definition at line 173 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::pSatMax |
Max value for parameterized saturation
Definition at line 175 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
uint8_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::mode |
mode for input feature map in Circular or Linear mode in B matrix
Definition at line 178 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int32_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::col |
column number of input feature map
Definition at line 180 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.
int8_t MMALIB_CNN_convolveBias_row_ixX_ixX_oxX_InitArgs::packetizeMode |
packetizeMode is 0 then weights are not packetized. When 1 then they are packetized
Definition at line 184 of file MMALIB_CNN_convolveBias_row_ixX_ixX_oxX.h.