Intel® Math Kernel Library 2019 Developer Reference - C
Creates propagation operations for rectified linear neuron activation 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.
dnnError_t dnnReLUCreateForward_F32 (dnnPrimitive_t *pRelu, dnnPrimitiveAttributes_t attributes, const dnnLayout_t dataLayout, float negativeSlope);
dnnError_t dnnReLUCreateBackward_F32 (dnnPrimitive_t *pRelu, dnnPrimitiveAttributes_t attributes, const dnnLayout_t diffLayout, const dnnLayout_t dataLayout, float negativeSlope);
dnnError_t dnnReLUCreateForward_F64 (dnnPrimitive_t *pRelu, dnnPrimitiveAttributes_t attributes, const dnnLayout_t dataLayout, double negativeSlope);
dnnError_t dnnReLUCreateBackward_F64 (dnnPrimitive_t *pRelu, dnnPrimitiveAttributes_t attributes, const dnnLayout_t diffLayout, const dnnLayout_t dataLayout, double negativeSlope);
attributes |
The set of attributes for the primitive. |
dataLayout |
The layout of the input. |
diffLayout |
The layout of the destination diff. |
negativeSlope |
The negative slope. |
pRelu |
Pointer to the primitive to create:
|
Each dnnReLUCreate function creates a forward or backward propagation operation for batch rectified linear neuron activation (ReLU). The ReLU operation is defined as:
dst[x] = max(src[x], 0) + negativeSlope*min(src[x], 0).