Public Member Functions
	TrainingConfig ()

	TrainingConfig (size_t numWorkers, size_t updateInterval, size_t targetNetworkSyncInterval, size_t stepLimit, size_t explorationSteps, double stepSize, double discount, double gradientLimit, bool doubleQLearning, bool noisyQLearning, bool isCategorical, size_t atomSize, double vMin, double vMax, double rho)

size_t	AtomSize () const
	Get the number of atoms. More...

size_t &	AtomSize ()
	Modify the number of atoms. More...

double	Discount () const
	Get the discount rate for future reward. More...

double &	Discount ()
	Modify the discount rate for future reward. More...

bool	DoubleQLearning () const
	Get the indicator of double q-learning. More...

bool &	DoubleQLearning ()
	Modify the indicator of double q-learning. More...

size_t	ExplorationSteps () const
	Get the exploration steps. More...

size_t &	ExplorationSteps ()
	Modify the exploration steps. More...

double	GradientLimit () const
	Get the limit of update gradient. More...

double &	GradientLimit ()
	Modify the limit of update gradient. More...

bool	IsCategorical () const
	Get the indicator of categorical q-learning. More...

bool &	IsCategorical ()
	Modify the indicator of categorical q-learning. More...

bool	NoisyQLearning () const
	Get the indicator of noisy q-learning. More...

bool &	NoisyQLearning ()
	Modify the indicator of double q-learning. More...

size_t	NumWorkers () const
	Get the amount of workers. More...

size_t &	NumWorkers ()
	Modify the amount of workers. More...

double	Rho () const
	Get the rho value for sac. More...

double &	Rho ()
	Modify the rho value for sac. More...

size_t	StepLimit () const
	Get the maximum steps of each episode. More...

size_t &	StepLimit ()
	Modify the maximum steps of each episode. More...

double	StepSize () const
	Get the step size of the optimizer. More...

double &	StepSize ()
	Modify the step size of the optimizer. More...

size_t	TargetNetworkSyncInterval () const
	Get the interval for syncing target network. More...

size_t &	TargetNetworkSyncInterval ()
	Modify the interval for syncing target network. More...

size_t	UpdateInterval () const
	Get the update interval. More...

size_t &	UpdateInterval ()
	Modify the update interval. More...

double	VMax () const
	Get the maximum value for support. More...

double &	VMax ()
	Modify the maximum value for support. More...

double	VMin () const
	Get the minimum value for support. More...

double &	VMin ()
	Modify the minimum value for support. More...

Detailed Description

Definition at line 19 of file training_config.hpp.

Constructor & Destructor Documentation

◆ TrainingConfig() [1/2]

TrainingConfig ( )

inline

Definition at line 22 of file training_config.hpp.

◆ TrainingConfig() [2/2]

TrainingConfig	(	size_t	numWorkers,
		size_t	updateInterval,
		size_t	targetNetworkSyncInterval,
		size_t	stepLimit,
		size_t	explorationSteps,
		double	stepSize,
		double	discount,
		double	gradientLimit,
		bool	doubleQLearning,
		bool	noisyQLearning,
		bool	isCategorical,
		size_t	atomSize,
		double	vMin,
		double	vMax,
		double	rho
	)

inline

Definition at line 40 of file training_config.hpp.

Member Function Documentation

◆ AtomSize() [1/2]

size_t AtomSize ( ) const

inline

Get the number of atoms.

Definition at line 133 of file training_config.hpp.

◆ AtomSize() [2/2]

size_t& AtomSize ( )

inline

Modify the number of atoms.

Definition at line 135 of file training_config.hpp.

◆ Discount() [1/2]

double Discount ( ) const

inline

Get the discount rate for future reward.

Definition at line 108 of file training_config.hpp.

Referenced by NStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), OneStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), and OneStepSarsaWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step().

◆ Discount() [2/2]

double& Discount ( )

inline

Modify the discount rate for future reward.

Definition at line 110 of file training_config.hpp.

◆ DoubleQLearning() [1/2]

bool DoubleQLearning ( ) const

inline

Get the indicator of double q-learning.

Definition at line 118 of file training_config.hpp.

◆ DoubleQLearning() [2/2]

bool& DoubleQLearning ( )

inline

Modify the indicator of double q-learning.

Definition at line 120 of file training_config.hpp.

◆ ExplorationSteps() [1/2]

size_t ExplorationSteps ( ) const

inline

Get the exploration steps.

Definition at line 98 of file training_config.hpp.

◆ ExplorationSteps() [2/2]

size_t& ExplorationSteps ( )

inline

Modify the exploration steps.

Definition at line 100 of file training_config.hpp.

◆ GradientLimit() [1/2]

double GradientLimit ( ) const

inline

Get the limit of update gradient.

Definition at line 113 of file training_config.hpp.

Referenced by NStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), OneStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), and OneStepSarsaWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step().

◆ GradientLimit() [2/2]

double& GradientLimit ( )

inline

Modify the limit of update gradient.

Definition at line 115 of file training_config.hpp.

◆ IsCategorical() [1/2]

bool IsCategorical ( ) const

inline

Get the indicator of categorical q-learning.

Definition at line 128 of file training_config.hpp.

◆ IsCategorical() [2/2]

bool& IsCategorical ( )

inline

Modify the indicator of categorical q-learning.

Definition at line 130 of file training_config.hpp.

◆ NoisyQLearning() [1/2]

bool NoisyQLearning ( ) const

inline

Get the indicator of noisy q-learning.

Definition at line 123 of file training_config.hpp.

◆ NoisyQLearning() [2/2]

bool& NoisyQLearning ( )

inline

Modify the indicator of double q-learning.

Definition at line 125 of file training_config.hpp.

◆ NumWorkers() [1/2]

size_t NumWorkers ( ) const

inline

Get the amount of workers.

Definition at line 74 of file training_config.hpp.

◆ NumWorkers() [2/2]

size_t& NumWorkers ( )

inline

Modify the amount of workers.

Definition at line 76 of file training_config.hpp.

◆ Rho() [1/2]

double Rho ( ) const

inline

Get the rho value for sac.

Definition at line 148 of file training_config.hpp.

◆ Rho() [2/2]

double& Rho ( )

inline

Modify the rho value for sac.

Definition at line 150 of file training_config.hpp.

◆ StepLimit() [1/2]

size_t StepLimit ( ) const

inline

Get the maximum steps of each episode.

Definition at line 90 of file training_config.hpp.

Referenced by NStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), OneStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), and OneStepSarsaWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step().

◆ StepLimit() [2/2]

size_t& StepLimit ( )

inline

Modify the maximum steps of each episode.

Setting it to 0 means no limit.

Definition at line 95 of file training_config.hpp.

◆ StepSize() [1/2]

double StepSize ( ) const

inline

Get the step size of the optimizer.

Definition at line 103 of file training_config.hpp.

Referenced by NStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), OneStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), and OneStepSarsaWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step().

◆ StepSize() [2/2]

double& StepSize ( )

inline

Modify the step size of the optimizer.

Definition at line 105 of file training_config.hpp.

◆ TargetNetworkSyncInterval() [1/2]

size_t TargetNetworkSyncInterval ( ) const

inline

Get the interval for syncing target network.

Definition at line 84 of file training_config.hpp.

Referenced by NStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), OneStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), and OneStepSarsaWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step().

◆ TargetNetworkSyncInterval() [2/2]

size_t& TargetNetworkSyncInterval ( )

inline

Modify the interval for syncing target network.

Definition at line 87 of file training_config.hpp.

◆ UpdateInterval() [1/2]

size_t UpdateInterval ( ) const

inline

Get the update interval.

Definition at line 79 of file training_config.hpp.

Referenced by NStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), OneStepQLearningWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step(), and OneStepSarsaWorker< EnvironmentType, NetworkType, UpdaterType, PolicyType >::Step().

◆ UpdateInterval() [2/2]

size_t& UpdateInterval ( )

inline

Modify the update interval.

Definition at line 81 of file training_config.hpp.

◆ VMax() [1/2]

double VMax ( ) const

inline

Get the maximum value for support.

Definition at line 143 of file training_config.hpp.

◆ VMax() [2/2]

double& VMax ( )

inline

Modify the maximum value for support.

Definition at line 145 of file training_config.hpp.

◆ VMin() [1/2]

double VMin ( ) const

inline

Get the minimum value for support.

Definition at line 138 of file training_config.hpp.

◆ VMin() [2/2]

double& VMin ( )

inline

Modify the minimum value for support.

Definition at line 140 of file training_config.hpp.

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

/home/ryan/src/mlpack.org/_src/mlpack-git/src/mlpack/methods/reinforcement_learning/training_config.hpp

Public Member Functions

Detailed Description

Constructor & Destructor Documentation

◆ TrainingConfig() [1/2]

◆ TrainingConfig() [2/2]

Member Function Documentation

◆ AtomSize() [1/2]

◆ AtomSize() [2/2]

◆ Discount() [1/2]

◆ Discount() [2/2]

◆ DoubleQLearning() [1/2]

◆ DoubleQLearning() [2/2]

◆ ExplorationSteps() [1/2]

◆ ExplorationSteps() [2/2]

◆ GradientLimit() [1/2]

◆ GradientLimit() [2/2]

◆ IsCategorical() [1/2]

◆ IsCategorical() [2/2]

◆ NoisyQLearning() [1/2]

◆ NoisyQLearning() [2/2]

◆ NumWorkers() [1/2]

◆ NumWorkers() [2/2]

◆ Rho() [1/2]

◆ Rho() [2/2]

◆ StepLimit() [1/2]

◆ StepLimit() [2/2]

◆ StepSize() [1/2]

◆ StepSize() [2/2]

◆ TargetNetworkSyncInterval() [1/2]

◆ TargetNetworkSyncInterval() [2/2]

◆ UpdateInterval() [1/2]

◆ UpdateInterval() [2/2]

◆ VMax() [1/2]

◆ VMax() [2/2]

◆ VMin() [1/2]

◆ VMin() [2/2]