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

Binary Cross Entropy with Logits Loss. More...

#include <losses.hpp>

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

Collaboration diagram for dl::loss::BCEWithLogitsLoss< T >:

Public Member Functions
	BCEWithLogitsLoss (const std::string &reduction="mean")
	Constructor.
Variable< T >	forward (const Variable< T > &predictions, const Variable< T > &targets) override
	Forward pass: compute BCE loss from logits.
Public Member Functions inherited from dl::loss::AutogradLoss< T >
virtual	~AutogradLoss ()=default
Variable< T >	operator() (const Variable< T > &predictions, const Variable< T > &targets)
	Convenience operator for computing loss.

Detailed Description

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

Binary Cross Entropy with Logits Loss.

Combines sigmoid and BCE for numerical stability. More stable than applying sigmoid then BCE separately.

Definition at line 148 of file losses.hpp.

Constructor & Destructor Documentation

BCEWithLogitsLoss()

template<typename T >

dl::loss::BCEWithLogitsLoss< T >::BCEWithLogitsLoss ( const std::string &  reduction = "mean" )

inlineexplicit

Constructor.

Parameters

reduction

Type of reduction ('mean', 'sum', 'none')

Definition at line 154 of file losses.hpp.

Member Function Documentation

forward()

template<typename T >

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

overridevirtual

Forward pass: compute BCE loss from logits.

Parameters

predictions	Raw logits (before sigmoid)
targets	Binary target values (0 or 1)

Returns

BCE loss

Implements dl::loss::AutogradLoss< T >.

Definition at line 121 of file losses.cpp.

---

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

• include/loss/losses.hpp
• src/loss/losses.cpp