agent/baseline_agent.py [122:144]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
                lr=decoder_lr,
                weight_decay=decoder_weight_lambda
            )
            # optimizer for critic encoder for reconstruction loss
            self.encoder_optimizer = torch.optim.Adam(
                self.critic.encoder.parameters(), lr=encoder_lr
            )

        # optimizers
        self.actor_optimizer = torch.optim.Adam(
            self.actor.parameters(), lr=actor_lr, betas=(actor_beta, 0.999)
        )

        self.critic_optimizer = torch.optim.Adam(
            self.critic.parameters(), lr=critic_lr, betas=(critic_beta, 0.999)
        )

        self.log_alpha_optimizer = torch.optim.Adam(
            [self.log_alpha], lr=alpha_lr, betas=(alpha_beta, 0.999)
        )

        self.train()
        self.critic_target.train()
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



agent/deepmdp_agent.py [116:139]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            lr=decoder_lr,
            weight_decay=decoder_weight_lambda
        )

        # optimizer for critic encoder for reconstruction loss
        self.encoder_optimizer = torch.optim.Adam(
            self.critic.encoder.parameters(), lr=encoder_lr
        )

        # optimizers
        self.actor_optimizer = torch.optim.Adam(
            self.actor.parameters(), lr=actor_lr, betas=(actor_beta, 0.999)
        )

        self.critic_optimizer = torch.optim.Adam(
            self.critic.parameters(), lr=critic_lr, betas=(critic_beta, 0.999)
        )

        self.log_alpha_optimizer = torch.optim.Adam(
            [self.log_alpha], lr=alpha_lr, betas=(alpha_beta, 0.999)
        )

        self.train()
        self.critic_target.train()
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -