nest::WeightOptimizerCommonPropertiesAdam Class Reference

Class implementing common properties of an Adam weight optimizer model. More...

#include <weight_optimizer.h>

Inheritance diagram for nest::WeightOptimizerCommonPropertiesAdam:

Collaboration diagram for nest::WeightOptimizerCommonPropertiesAdam:

Public Member Functions
	WeightOptimizerCommonPropertiesAdam ()
	Default constructor.
	WeightOptimizerCommonPropertiesAdam (const WeightOptimizerCommonPropertiesAdam &)=default
	Copy constructor.
WeightOptimizerCommonPropertiesAdam &	operator= (const WeightOptimizerCommonPropertiesAdam &)=delete
	Assignment operator.
WeightOptimizerCommonProperties *	clone () const override
	Clone constructor.
WeightOptimizer *	get_optimizer () const override
	Get optimizer.
void	get_status (Dictionary &d) const override
	Get parameter dictionary.
void	set_status (const Dictionary &d) override
	Update parameters in parameter dictionary.
std::string	get_name () const override
	Get optimizer name.
Public Member Functions inherited from nest::WeightOptimizerCommonProperties
	WeightOptimizerCommonProperties ()
	Default constructor.
virtual	~WeightOptimizerCommonProperties ()
	Destructor.
	WeightOptimizerCommonProperties (const WeightOptimizerCommonProperties &)
	Copy constructor.
WeightOptimizer &	operator= (const WeightOptimizer &)=delete
	Assignment operator.
double	get_Wmin () const
	Get minimal value for synaptic weight.
double	get_Wmax () const
	Get maximal value for synaptic weight.

Private Attributes
double	beta_1_
	Exponential decay rate for first moment estimate.
double	beta_2_
	Exponential decay rate for second moment estimate.
double	epsilon_
	Small constant for numerical stability.

Friends
class	WeightOptimizerAdam
	Friend class for Adam weight optimizer model.

Additional Inherited Members
Public Attributes inherited from nest::WeightOptimizerCommonProperties
size_t	batch_size_
	Size of an optimization batch.
double	eta_
	Common learning rate for all synapses.
double	eta_first_change_
	First non-default learning rate.
long	eta_change_count_
	Count of learning rate changes so far in the simulation to identify the first change.
double	Wmin_
	Minimal value for synaptic weight.
double	Wmax_
	Maximal value for synaptic weight.
bool	optimize_each_step_
	If true, optimize each step, else once per spike.

Detailed Description

Class implementing common properties of an Adam weight optimizer model.

Constructor & Destructor Documentation

WeightOptimizerCommonPropertiesAdam() [1/2]

nest::WeightOptimizerCommonPropertiesAdam::WeightOptimizerCommonPropertiesAdam

(

)

Default constructor.

Referenced by clone().

Here is the caller graph for this function:

WeightOptimizerCommonPropertiesAdam() [2/2]

nest::WeightOptimizerCommonPropertiesAdam::WeightOptimizerCommonPropertiesAdam ( const WeightOptimizerCommonPropertiesAdam &  )

default

Copy constructor.

Member Function Documentation

clone()

WeightOptimizerCommonProperties * nest::WeightOptimizerCommonPropertiesAdam::clone ( ) const

overridevirtual

Clone constructor.

Implements nest::WeightOptimizerCommonProperties.

References WeightOptimizerCommonPropertiesAdam().

Here is the call graph for this function:

get_name()

std::string nest::WeightOptimizerCommonPropertiesAdam::get_name ( ) const

inlineoverridevirtual

Get optimizer name.

Implements nest::WeightOptimizerCommonProperties.

get_optimizer()

WeightOptimizer * nest::WeightOptimizerCommonPropertiesAdam::get_optimizer ( ) const

overridevirtual

Get optimizer.

Implements nest::WeightOptimizerCommonProperties.

References WeightOptimizerAdam.

get_status()

void nest::WeightOptimizerCommonPropertiesAdam::get_status ( Dictionary &  d ) const

overridevirtual

Get parameter dictionary.

Reimplemented from nest::WeightOptimizerCommonProperties.

References nest::names::beta_1(), beta_1_, nest::names::beta_2(), beta_2_, nest::names::epsilon(), epsilon_, and nest::WeightOptimizerCommonProperties::get_status().

Here is the call graph for this function:

operator=()

WeightOptimizerCommonPropertiesAdam & nest::WeightOptimizerCommonPropertiesAdam::operator= ( const WeightOptimizerCommonPropertiesAdam &  )

delete

Assignment operator.

set_status()

void nest::WeightOptimizerCommonPropertiesAdam::set_status ( const Dictionary &  d )

overridevirtual

Update parameters in parameter dictionary.

Reimplemented from nest::WeightOptimizerCommonProperties.

References nest::names::beta_1(), beta_1_, nest::names::beta_2(), beta_2_, nest::names::epsilon(), epsilon_, and nest::WeightOptimizerCommonProperties::set_status().

Here is the call graph for this function:

Friends And Related Symbol Documentation

WeightOptimizerAdam

friend class WeightOptimizerAdam

friend

Friend class for Adam weight optimizer model.

Referenced by get_optimizer().

Member Data Documentation

beta_1_

double nest::WeightOptimizerCommonPropertiesAdam::beta_1_

private

Exponential decay rate for first moment estimate.

Referenced by get_status(), nest::WeightOptimizerAdam::optimize_(), and set_status().

beta_2_

double nest::WeightOptimizerCommonPropertiesAdam::beta_2_

private

Exponential decay rate for second moment estimate.

Referenced by get_status(), nest::WeightOptimizerAdam::optimize_(), and set_status().

epsilon_

double nest::WeightOptimizerCommonPropertiesAdam::epsilon_

private

Small constant for numerical stability.

Referenced by get_status(), nest::WeightOptimizerAdam::optimize_(), and set_status().

---

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

• weight_optimizer.h
• weight_optimizer.cpp