Module border::agent [−][src]
Expand description
Agents of reinforcement learning.
In this crate, an agent is defined as an trainable policy. Currently, agents are implemented with tch-rs.
Re-exports
pub use self::tch::dqn::DQN; | |
pub use self::tch::replay_buffer::ReplayBuffer; | |
pub use self::tch::replay_buffer::TchBuffer; | |
pub use base::CriticLoss; | |
pub use base::OptInterval; | |
pub use base::OptIntervalCounter; |
Modules
base | Utilities used by agents. |
tch | Agents implemented based on tch-rs. |