NAME

MPSCNNLogSoftMaxGradient

SYNOPSIS

#import <MPSCNNSoftMax.h>

Inherits MPSCNNGradientKernel.

Instance Methods

(nonnull instancetype) - initWithDevice:
(nullable instancetype) - initWithCoder:device:

Detailed Description

This depends on Metal.framework The logSoftMax gradient filter calculates the gradient to be backpropagated. The logSoftMax gradient just as the log softMax filter, is applied across feature channels and at all spatial locations. It computes the gradient for a given output generated by the corresponding logSoftMax (i.e. MPSCNNLogSoftMax) layer and the gradient computed by the previous layer in the back-propagation pass. For each feature channel per pixel in an image in a feature map, the logSoftMax gradient filter computes the following: result gradient channel in pixel outputGradient(x,y,k) = inputGradient(x,y,k) - exp(logSoftMax(x,y,k)) * sum(inputGradient(x,y,0) ... inputGradient(x,y,N-1)) where N is the number of feature channels

The incoming gradient is the primary source. The original output of corresponding logSoftMax is the secondary source.

Method Documentation

- (nullable instancetype) initWithCoder: (NSCoder *__nonnull) aDecoder(nonnull id< MTLDevice >) device

NSSecureCoding compatability While the standard NSSecureCoding/NSCoding method -initWithCoder: should work, since the file can't know which device your data is allocated on, we have to guess and may guess incorrectly. To avoid that problem, use initWithCoder:device instead.

Parameters:

aDecoder The NSCoder subclass with your serialized MPSKernel
device The MTLDevice on which to make the MPSKernel

Returns:

A new MPSKernel object, or nil if failure.

Reimplemented from MPSCNNGradientKernel.

- (nonnull instancetype) initWithDevice: (nonnull id< MTLDevice >) device

Initializes a MPSCNNLogSoftMaxGradient function

Parameters:

device The MTLDevice on which this MPSCNNLogSoftMaxGradient filter will be used