Struct RLOptimizerConfig 

pub struct RLOptimizerConfig<T: Float + Debug + Send + Sync + 'static> {

    pub policy_lr: T,
    pub value_lr: T,
    pub discount_factor: T,
    pub gae_lambda: T,
    pub clip_epsilon: T,
    pub entropy_coeff: T,
    pub value_loss_coeff: T,
    pub max_grad_norm: T,
    pub n_epochs: usize,
    pub mini_batchsize: usize,
    pub trust_region_config: Option<TrustRegionConfig<T>>,
    pub use_natural_gradients: bool,
    pub fisher_approximation: FisherApproximationMethod,
}

Reinforcement Learning optimization configuration

Fields

policy_lr: T

Policy learning rate

value_lr: T

Value function learning rate

discount_factor: T

Discount factor (gamma)

gae_lambda: T

GAE lambda parameter

clip_epsilon: T

Clipping parameter for PPO

entropy_coeff: T

Entropy regularization coefficient

value_loss_coeff: T

Value function loss coefficient

max_grad_norm: T

Maximum gradient norm for clipping

n_epochs: usize

Number of optimization epochs per update

mini_batchsize: usize

Mini-batch size for optimization

trust_region_config: Option<TrustRegionConfig<T>>

Trust region methods configuration

use_natural_gradients: bool

Enable natural policy gradients

fisher_approximation: FisherApproximationMethod

Fisher information matrix approximation method

Trait Implementations

impl<T: Clone + Float + Debug + Send + Sync + 'static> Clone for RLOptimizerConfig<T>

fn clone(&self) -> RLOptimizerConfig<T>

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl<T: Debug + Float + Debug + Send + Sync + 'static> Debug for RLOptimizerConfig<T>

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<T: Float + Debug + Send + Sync + 'static> Default for RLOptimizerConfig<T>

fn default() -> Self

Returns the “default value” for a type.