Creates propagation operations for inner product layers. Note: The Deep Neural Network (DNN) component in Intel MKL is deprecated and will be removed in a future release. You can continue to use optimized functions for deep neural networks through Intel Math Kernel Library for Deep Neural Networks.

Syntax

dnnError_t dnnInnerProductCreateForward_F32 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t srcSize[], size_t outputChannels);

dnnError_t dnnInnerProductCreateForwardBias_F32 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t srcSize[], size_t outputChannels);

dnnError_t dnnInnerProductCreateBackwardData_F32 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t srcSize[], size_t outputChannels);

dnnError_t dnnInnerProductCreateBackwardFilter_F32 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t srcSize[], size_t outputChannels);

dnnError_t dnnInnerProductCreateBackwardBias_F32 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t dstSize[]);

dnnError_t dnnInnerProductCreateForward_F64 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t srcSize[], size_t outputChannels);

dnnError_t dnnInnerProductCreateForwardBias_F64 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t srcSize[], size_t outputChannels);

dnnError_t dnnInnerProductCreateBackwardData_F64 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t srcSize[], size_t outputChannels);

dnnError_t dnnInnerProductCreateBackwardFilter_F64 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t srcSize[], size_t outputChannels);

dnnError_t dnnInnerProductCreateBackwardBias_F64 (dnnPrimitive_t *pInnerProduct, dnnPrimitiveAttributes_t attributes, size_t dimentions, const size_t dstSize[]);

Include Files

  • mkl.h

Input Parameters

attributes

The set of attributes for the primitive.

dimension

The number of dimensions in the input and output.

srcSize

srcSize[i] is the size of the input along the dimension i.

dstSize

dstSize[i] is the size of the output along the dimension i.

outputChannels

The number of output channels.

Output Parameters

pInnerProduct

Pointer to the primitive to create:

dnnInnerProductCreateForward

Forward without bias

dnnInnerProductCreateForwardBias

Forward with bias

dnnInnerProductCreateBackwardData

Backward with respect to data

dnnInnerProductCreateBackwardFilter

Backward with respect to filter

dnnInnerProductCreateBackwardBias

Backward with respect to bias

Description

Each dnnInnerProductCreate function creates a forward or backward propagation operation for the inner product. The inner product operation is defined as:





Для получения подробной информации о возможностях оптимизации компилятора обратитесь к нашему Уведомлению об оптимизации.
Выберите цвет фиксируемой кнопки: 
Orange (only for download buttons)