24.02.1
|
#include <KernelDescriptors.h>
Public Member Functions | |
InstanceNormalizationLayerKernelInfo () | |
Default constructor. More... | |
InstanceNormalizationLayerKernelInfo (float gamma, float beta, float epsilon, bool use_mixed_precision) | |
Constructor. More... | |
Data Fields | |
float | gamma |
The scale scalar value applied to the normalized tensor. More... | |
float | beta |
The offset scalar value applied to the normalized tensor. More... | |
float | epsilon |
Lower bound value for the normalization. More... | |
bool | use_mixed_precision |
Use mixed precision in case of FP16 execution. More... | |
Definition at line 153 of file KernelDescriptors.h.
|
inline |
Default constructor.
Definition at line 156 of file KernelDescriptors.h.
|
inline |
Constructor.
[in] | gamma | The scale scalar value applied to the normalized tensor. |
[in] | beta | The offset scalar value applied to the normalized tensor |
[in] | epsilon | Lower bound value for the normalization. |
[in] | use_mixed_precision | Use mixed precision in case of FP16 execution. |
Definition at line 166 of file KernelDescriptors.h.
float beta |
The offset scalar value applied to the normalized tensor.
Defaults to 0.0
Definition at line 172 of file KernelDescriptors.h.
float epsilon |
Lower bound value for the normalization.
Defaults to 1e-12
Definition at line 173 of file KernelDescriptors.h.
float gamma |
The scale scalar value applied to the normalized tensor.
Defaults to 1.0
Definition at line 171 of file KernelDescriptors.h.
bool use_mixed_precision |
Use mixed precision in case of FP16 execution.
Defaults to true
Definition at line 174 of file KernelDescriptors.h.