Java* API Reference for Intel® Data Analytics Acceleration Library 2019 Update 5
Identifiers of input objects for the backward batch normalization layer and results for the forward batch normalization layer. More...
BatchNormalizationLayerDataId | ( | int | value | ) |
Constructs the input object identifier using the provided value
value | Value corresponding to the input object identifier |
Static Public Attributes | |
static final BatchNormalizationLayerDataId | auxData = new BatchNormalizationLayerDataId(auxDataId) |
static final BatchNormalizationLayerDataId | auxWeights = new BatchNormalizationLayerDataId(auxWeightsId) |
static final BatchNormalizationLayerDataId | auxMean = new BatchNormalizationLayerDataId(auxMeanId) |
static final BatchNormalizationLayerDataId | auxStandardDeviation = new BatchNormalizationLayerDataId(auxStandardDeviationId) |
static final BatchNormalizationLayerDataId | auxPopulationMean = new BatchNormalizationLayerDataId(auxPopulationMeanId) |
static final BatchNormalizationLayerDataId | auxPopulationVariance = new BatchNormalizationLayerDataId(auxPopulationVarianceId) |
int getValue | ( | ) |
Returns the value corresponding to the input object identifier
|
static |
p-dimensional tensor that stores forward batch normalization layer input data
|
static |
1-dimensional tensor of size that stores mini-batch mean
|
static |
1-dimensional tensor of size that stores resulting population mean
|
static |
1-dimensional tensor of size that stores resulting population variance
|
static |
1-dimensional tensor of size that stores mini-batch standard deviation
|
static |
1-dimensional tensor of size that stores input weights for forward batch normalization layer
For more complete information about compiler optimizations, see our Optimization Notice.