AIFES_E_training_parameter_fnn_f32 Struct Reference

Parameters for Training. More...

#include <aifes_express_f32_fnn.h>

Data Fields
AIFES_E_loss	loss
	Loss selection (AIFES_E_loss)
AIFES_E_optimizer	optimizer
	Optimizer selection (AIFES_E_optimizer)
float	learn_rate
	Learning rate for training (For all optimizers)
float	sgd_momentum
	Optional momentum for SGD (Value 0.0f means Momentum off)
uint32_t	batch_size
	Batch size (min: 1 -> max: Entire data set)
uint32_t	epochs
	Number of desired epochs (If early stopping is on, can be stopped before)
uint32_t	epochs_loss_print_interval
	Selection of the interval in which the loss is to be calculated and output via the print function.
void(*	loss_print_function )(float)
	Individual print function for the loss. More...
AIFES_E_early_stopping	early_stopping
	Switch early stopping on or off.
float	early_stopping_target_loss
	If early stopping is switched on, the target loss can be specified here. More...

Detailed Description

Parameters for Training.

The structure includes all parameters for training.

Example:

AIFES_E_training_parameter_fnn_f32  nn_train_config;
nn_train_config.optimizer = AIfES_E_adam;
nn_train_config.sgd_momentum = 0.0f;         // Only for SGD optimizer
nn_train_config.loss = AIfES_E_mse;
nn_train_config.learn_rate = 0.05f;
nn_train_config.batch_size = 4;
nn_train_config.epochs = 100;
nn_train_config.epochs_loss_print_interval = 10;
 
// Your individual print function
nn_train_config.loss_print_function = my_print_function;
 
//You can enable early stopping, so that learning is automatically stopped when a learning target is reached
nn_train_config.early_stopping = AIfES_E_early_stopping_on;
//Define your target loss
nn_train_config.early_stopping_target_loss = 0.004;

Field Documentation

early_stopping_target_loss

float early_stopping_target_loss

If early stopping is switched on, the target loss can be specified here.

It is only checked in the print interval

loss_print_function

void(* loss_print_function) (float)

Individual print function for the loss.

The loss is calculated only in the previously selected interval and output via the function to which the pointer points The function must have the following structure

void my_print_function(float loss)
{
   //E.g. a normal print output
   printf("Loss: %f\n", loss);
}

Parameters

tensor

The tensor to print

---

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

• aifes_express_f32_fnn.h