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) | |
| 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... | |
| size_t | NumWorkers () const |
| Get the amount of workers. More... | |
| size_t & | NumWorkers () |
| Modify the amount of workers. 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... | |
Detailed Description
Definition at line 19 of file training_config.hpp.
Constructor & Destructor Documentation
◆ TrainingConfig() [1/2]
|
inline |
Definition at line 22 of file training_config.hpp.
◆ TrainingConfig() [2/2]
|
inline |
Definition at line 31 of file training_config.hpp.
Member Function Documentation
◆ Discount() [1/2]
|
inline |
Get the discount rate for future reward.
Definition at line 87 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]
|
inline |
Modify the discount rate for future reward.
Definition at line 89 of file training_config.hpp.
◆ DoubleQLearning() [1/2]
|
inline |
Get the indicator of double q-learning.
Definition at line 97 of file training_config.hpp.
◆ DoubleQLearning() [2/2]
|
inline |
Modify the indicator of double q-learning.
Definition at line 99 of file training_config.hpp.
◆ ExplorationSteps() [1/2]
|
inline |
Get the exploration steps.
Definition at line 77 of file training_config.hpp.
◆ ExplorationSteps() [2/2]
|
inline |
Modify the exploration steps.
Definition at line 79 of file training_config.hpp.
◆ GradientLimit() [1/2]
|
inline |
Get the limit of update gradient.
Definition at line 92 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]
|
inline |
Modify the limit of update gradient.
Definition at line 94 of file training_config.hpp.
◆ NumWorkers() [1/2]
|
inline |
Get the amount of workers.
Definition at line 53 of file training_config.hpp.
◆ NumWorkers() [2/2]
|
inline |
Modify the amount of workers.
Definition at line 55 of file training_config.hpp.
◆ StepLimit() [1/2]
|
inline |
Get the maximum steps of each episode.
Definition at line 69 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]
|
inline |
Modify the maximum steps of each episode.
Setting it to 0 means no limit.
Definition at line 74 of file training_config.hpp.
◆ StepSize() [1/2]
|
inline |
Get the step size of the optimizer.
Definition at line 82 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]
|
inline |
Modify the step size of the optimizer.
Definition at line 84 of file training_config.hpp.
◆ TargetNetworkSyncInterval() [1/2]
|
inline |
Get the interval for syncing target network.
Definition at line 63 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]
|
inline |
Modify the interval for syncing target network.
Definition at line 66 of file training_config.hpp.
◆ UpdateInterval() [1/2]
|
inline |
Get the update interval.
Definition at line 58 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]
|
inline |
Modify the update interval.
Definition at line 60 of file training_config.hpp.
The documentation for this class was generated from the following file:
- /home/jenkins-mlpack/mlpack.org/_src/mlpack-3.2.1/src/mlpack/methods/reinforcement_learning/training_config.hpp