FRLlibPPOSettings

A struct to hold Proximal Policy Optimization(PPO) settings for an RLLib training script.

struct FRLlibPPOSettings : public FTrainingSettings

Methods

GenerateTrainingArgs

virtual void GenerateTrainingArgs(FScriptArgBuilder &ArgBuilder) const

Appends PPO-related CLI arguments to the training script builder.

Parameters

• ArgBuilder (FScriptArgBuilder)

---

FRLlibPPOSettings

virtual ~FRLlibPPOSettings()

Attributes

GAELambda

float GAELambda = 0.95

GAE lambda for advantage estimation.

---

ClipParam

float ClipParam = 0.2

PPO surrogate clipping range.

---

bUseGAE

bool bUseGAE = true

Whether to use generalized advantage estimation.

Source: Source/ScholaTraining/Public/TrainingSettings/Ray/Algorithms/RLlibPPOSettings.h