Navigation
API > API/Plugins > API/Plugins/LearningTraining
Inheritance Hierarchy
- IPPOTrainer
- FSharedMemoryPPOTrainer
References
| Module | LearningTraining |
| Header | /Engine/Plugins/Experimental/LearningAgents/Source/LearningTraining/Public/LearningPPOTrainer.h |
| Include | #include "LearningPPOTrainer.h" |
Syntax
struct FSharedMemoryPPOTrainer : public UE::Learning::IPPOTrainer
Remarks
Trainer that uses shared memory and a Python sub-process to perform training
This trainer is the most simple and efficient when training the policy on the same computer that experience is being gathered on.
Constructors
| Type | Name | Description | |
|---|---|---|---|
FSharedMemoryPPOTrainer
(
const FString& TaskName, |
Creates a new Shared Memory PPO trainer |
Destructors
| Type | Name | Description | |
|---|---|---|---|
Overridden from IPPOTrainer
| Type | Name | Description | |
|---|---|---|---|
| ETrainerResponse | RecvCritic
(
FNeuralNetwork& OutNetwork, |
Wait for the trainer to push an updated critic network. | |
| ETrainerResponse | RecvPolicy
(
FNeuralNetwork& OutNetwork, |
Wait for the trainer to push an updated policy network. | |
| ETrainerResponse | SendCritic
(
const FNeuralNetwork& Network, |
Wait for the trainer to be ready and push the current critic network. | |
| ETrainerResponse | SendExperience
(
const FReplayBuffer& ReplayBuffer, |
Wait for the trainer to be ready and push new experience. | |
| ETrainerResponse | SendPolicy
(
const FNeuralNetwork& Network, |
Wait for the trainer to be ready and push the current policy network. | |
| ETrainerResponse | SendStop
(
const float Timeout |
Signal for the trainer to stop. | |
| void | Terminate () |
Terminate the trainer immediately. | |
| ETrainerResponse | Wait
(
const float Timeout |
Wait for the trainer to finish. |