opennn::EpochStats Struct Reference

Aggregate metrics produced for a single training or evaluation epoch. More...

#include <optimizer.h>

Public Attributes
float	error = 0.0f
	Mean loss over the epoch.
float	accuracy = 0.0f
	Mean classification accuracy over the epoch (0 if not classification).

Detailed Description

Aggregate metrics produced for a single training or evaluation epoch.

Member Data Documentation

accuracy

float opennn::EpochStats::accuracy = 0.0f

Mean classification accuracy over the epoch (0 if not classification).

error

float opennn::EpochStats::error = 0.0f

Mean loss over the epoch.