mlask::Layer Class Reference

Base class for all layers in the neural network. More...

#include <Layer.hpp>

Inheritance diagram for mlask::Layer:

Public Member Functions
std::size_t	getIn () const
std::size_t	getOut () const
virtual	~Layer ()=default
virtual vectorOut	forward (vectorIn)=0
	Defines a way to move forward in a neural network.
virtual vectorIn	backward (vectorOut)=0
	Defines a way to backpropagate error in backropagation algorithm.
virtual void	fit (float_t learning_rate)=0
	Describes how the layer 'learns', meaning it defines how layer updates itself.
virtual bool	tryConvertToONNX (onnx::GraphProto *graph, std::string input, std::string output) const
	Converts the layer to ONNX format. If the layer cannot be converted, it should return false.
virtual std::string	str () const
	Returns a string representation of the layer.

Protected Attributes
std::size_t	in_
std::size_t	out_

Detailed Description

Base class for all layers in the neural network.

Constructor & Destructor Documentation

~Layer()

virtual mlask::Layer::~Layer ( )

virtualdefault

Member Function Documentation

backward()

virtual vectorIn mlask::Layer::backward ( vectorOut  )

pure virtual

Defines a way to backpropagate error in backropagation algorithm.

Implemented in mlask::LambdaActivationFunction, mlask::FullyConnectedLayer< in, out >, mlask::SoftMax, and mlask::ActivationFunction.

fit()

virtual void mlask::Layer::fit ( float_t  learning_rate )

pure virtual

Describes how the layer 'learns', meaning it defines how layer updates itself.

Implemented in mlask::SoftMax, mlask::ActivationFunction, mlask::LambdaActivationFunction, and mlask::FullyConnectedLayer< in, out >.

forward()

virtual vectorOut mlask::Layer::forward ( vectorIn  )

pure virtual

Defines a way to move forward in a neural network.

Implemented in mlask::LambdaActivationFunction, mlask::FullyConnectedLayer< in, out >, mlask::SoftMax, and mlask::ActivationFunction.

getIn()

std::size_t mlask::Layer::getIn ( ) const

inline

getOut()

std::size_t mlask::Layer::getOut ( ) const

inline

str()

virtual std::string mlask::Layer::str ( ) const

inlinevirtual

Returns a string representation of the layer.

Reimplemented in mlask::LeakyRelu, mlask::Relu, mlask::Sigmoid, mlask::SoftMax, mlask::Tanh, mlask::ActivationFunction, and mlask::FullyConnectedLayer< in, out >.

tryConvertToONNX()

virtual bool mlask::Layer::tryConvertToONNX ( onnx::GraphProto *  graph, std::string  input, std::string  output  ) const

inlinevirtual

Converts the layer to ONNX format. If the layer cannot be converted, it should return false.

Reimplemented in mlask::LeakyRelu, mlask::Relu, mlask::Sigmoid, mlask::SoftMax, mlask::Tanh, and mlask::FullyConnectedLayer< in, out >.

Member Data Documentation

in_

std::size_t mlask::Layer::in_

protected

out_

std::size_t mlask::Layer::out_

protected

---

The documentation for this class was generated from the following file:

• include/Layer.hpp