MiniDNN::Callback Class Reference
#include <Callback.h>
Inheritance diagram for MiniDNN::Callback:
Public Member Functions | |
| virtual void | pre_training_batch (const Network *net, const Matrix &x, const Matrix &y) |
| virtual void | pre_training_batch (const Network *net, const Matrix &x, const IntegerVector &y) |
| virtual void | post_training_batch (const Network *net, const Matrix &x, const Matrix &y) |
| virtual void | post_training_batch (const Network *net, const Matrix &x, const IntegerVector &y) |
Public Attributes | |
| int | m_nbatch |
| int | m_batch_id |
| int | m_nepoch |
| int | m_epoch_id |
Detailed Description
The interface and default implementation of the callback function during model fitting. The purpose of this class is to allow users printing some messages in each epoch or mini-batch training, for example the time spent, the loss function values, etc.
This default implementation is a silent version of the callback function that basically does nothing. See the VerboseCallback class for a verbose version that prints the loss function value in each mini-batch.
Definition at line 28 of file Callback.h.
The documentation for this class was generated from the following file:
