Crate burn_rl 

Reinforcement learning building blocks for Burn.

• Environment: a simulation an agent acts in, returning a StepResult per action.
• Policy: maps observations to actions; PolicyLearner updates a policy from experience.
• TransitionBuffer: a replay buffer of Transitions sampled in batches.

Applications enable the rl feature of burn and use this crate as burn::rl. The training loop for agents lives in burn-train under its rl feature.

Re-exports

pub use environment::*;

pub use policy::*;

pub use transition_buffer::*;

Modules

environment

Module for implementing an environment.

policy

Module for implementing a policy.

transition_buffer

Transition buffer.