dl::loss::AutogradLoss< T > Class Template Reference

Base class for autograd-compatible loss functions. More...

#include <losses.hpp>

Inheritance diagram for dl::loss::AutogradLoss< T >:

Public Member Functions
virtual	~AutogradLoss ()=default
virtual Variable< T >	forward (const Variable< T > &predictions, const Variable< T > &targets)=0
	Compute loss between predictions and targets.
Variable< T >	operator() (const Variable< T > &predictions, const Variable< T > &targets)
	Convenience operator for computing loss.

Detailed Description

template<typename T>
class dl::loss::AutogradLoss< T >

Base class for autograd-compatible loss functions.

Definition at line 27 of file losses.hpp.

Constructor & Destructor Documentation

~AutogradLoss()

template<typename T >

virtual dl::loss::AutogradLoss< T >::~AutogradLoss ( )

virtualdefault

Member Function Documentation

forward()

template<typename T >

virtual Variable< T > dl::loss::AutogradLoss< T >::forward ( const Variable< T > &  predictions, const Variable< T > &  targets  )

pure virtual

Compute loss between predictions and targets.

Parameters

predictions	Model predictions
targets	Ground truth targets

Returns

Loss value as a Variable (scalar)

Implemented in dl::loss::MSELoss< T >, dl::loss::CrossEntropyLoss< T >, dl::loss::BCELoss< T >, dl::loss::BCEWithLogitsLoss< T >, dl::loss::HingeLoss< T >, and dl::loss::HuberLoss< T >.

operator()()

template<typename T >

Variable< T > dl::loss::AutogradLoss< T >::operator() ( const Variable< T > &  predictions, const Variable< T > &  targets  )

inline

Convenience operator for computing loss.

Definition at line 42 of file losses.hpp.

---

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

• include/loss/losses.hpp