Navigation
API > API/Plugins > API/Plugins/LearningTraining
References
| Module | LearningTraining |
| Header | /Engine/Plugins/Experimental/LearningAgents/Source/LearningTraining/Public/LearningPPOTrainer.h |
| Include | #include "LearningPPOTrainer.h" |
Syntax
struct FPPOTrainerNetworkSettings
Remarks
Settings for the networks used for training PPO. These settings must match the Neural Network objects passed to PPOTrainer::Train and the action noise must match the action noise used by the policy while gathering experience.
Variables
| Type | Name | Description | |
|---|---|---|---|
 |
EActivationFunction | CriticActivationFunction | Activation function to use on hidden layers of the critic network |
 |
int32 | CriticHiddenLayerSize | Number of neurons in each hidden layer of the critic network |
 |
int32 | CriticLayerNum | Total layers for critic network including input, hidden, and output layers |
 |
float | PolicyActionNoiseMax | Maximum action noise used by the policy |
 |
float | PolicyActionNoiseMin | Minimum action noise used by the policy |
 |
EActivationFunction | PolicyActivationFunction | Activation function to use on hidden layers of the policy network |
 |
int32 | PolicyHiddenLayerSize | Number of neurons in each hidden layer of the policy network |
 |
int32 | PolicyLayerNum | Total layers for policy network including input, hidden, and output layers |