Struct FSB3SACSettings

Struct FSB3SACSettings

• Defined in File SB3SACSettings.h

Inheritance Relationships

Base Type

• public FTrainingSettings (Struct FTrainingSettings)

struct FSB3SACSettings : public FTrainingSettings

Source: Source/Schola/Training/Public/TrainingSettings/StableBaselines/Algorithms/SB3SACSettings.h

Dependencies: FScriptArgBuilder

A struct to hold SAC settings for an SB3 training script.

Note

This is a partial implementation of the SAC settings, and is not exhaustive

---

Public Functions

Symbol	Details
GenerateTrainingArgs	Generate the training arguments for the script using the ArgBuilder.
~FSB3SACSettings	—

GenerateTrainingArgs

virtual void GenerateTrainingArgs(FScriptArgBuilder &ArgBuilder) const

Generate the training arguments for the script using the ArgBuilder.

Populates the ArgBuilder with training-specific command-line arguments.

Parameters

ArgBuilder – [inout] The builder to use to generate the arguments.

#	Direction	Name	Type	Description
1	—	ArgBuilder	FScriptArgBuilder &	The builder to use to generate the arguments.

Attributes: const, virtual

Source: Source/Schola/Training/Public/TrainingSettings/StableBaselines/Algorithms/SB3SACSettings.h

Implementation: Source/Schola/Training/Private/TrainingSettings/StableBaselines/Algorithms/SB3SACSettings.cpp

---

~FSB3SACSettings

virtual ~FSB3SACSettings()

Attributes: virtual

Source: Source/Schola/Training/Public/TrainingSettings/StableBaselines/Algorithms/SB3SACSettings.h

Implementation: Source/Schola/Training/Private/TrainingSettings/StableBaselines/Algorithms/SB3SACSettings.cpp

---

Public Members

Symbol	Details
LearningRate	The learning rate for the SAC algorithm.
BufferSize	The buffer size for the SAC algorithm.
LearningStarts	The number of steps to take before learning starts.
BatchSize	The batch size to use during gradient descent.
Tau	The Tau value for the SAC algorithm.
Gamma	The gamma value for the SAC algorithm.
TrainFreq	The frequency to update the target network, in steps.
GradientSteps	The number of gradient steps to take during training.
OptimizeMemoryUsage	Optimize memory usage.
LearnEntCoef	Should we learn the entropy coefficient during training.
InitialEntCoef	The initial entropy coefficient for the SAC algorithm.
TargetUpdateInterval	The interval at which we update the target network.
TargetEntropy	The target entropy for the SAC algorithm.
UseSDE	Use state dependent entropy noise.
SDESampleFreq	The frequency to sample the state dependent entropy noise.

---

LearningRate

float LearningRate = 0.0003

The learning rate for the SAC algorithm.

---

BufferSize

int BufferSize = 1000000

The buffer size for the SAC algorithm.

---

LearningStarts

int LearningStarts = 100

The number of steps to take before learning starts.

---

BatchSize

int BatchSize = 256

The batch size to use during gradient descent.

---

Tau

float Tau = 0.005

The Tau value for the SAC algorithm.

---

Gamma

float Gamma = 0.99

The gamma value for the SAC algorithm.

---

TrainFreq

int TrainFreq = 1

The frequency to update the target network, in steps.

---

GradientSteps

int GradientSteps = 1

The number of gradient steps to take during training.

---

OptimizeMemoryUsage

bool OptimizeMemoryUsage = false

Optimize memory usage.

---

LearnEntCoef

bool LearnEntCoef = true

Should we learn the entropy coefficient during training.

---

InitialEntCoef

float InitialEntCoef = 1.0

The initial entropy coefficient for the SAC algorithm.

---

TargetUpdateInterval

int TargetUpdateInterval = 1

The interval at which we update the target network.

---

TargetEntropy

FString TargetEntropy = "auto"

The target entropy for the SAC algorithm.

use auto to learn the target entropy

---

UseSDE

bool UseSDE = false

Use state dependent entropy noise.

---

SDESampleFreq

int SDESampleFreq = -1

The frequency to sample the state dependent entropy noise.