def prep_training(self, train_device):
self.q1.train()
self.q2.train()
to_device(self.q1, train_device)
to_device(self.q2, train_device)
to_device(self.tq1, train_device)
to_device(self.tq2, train_device)
to_device(self.log_alpha, train_device)
def prep_training(self, train_device):
self.pi.train()
self.q1.train()
self.q2.train()
to_device(self.pi, train_device)
to_device(self.q1, train_device)
to_device(self.q2, train_device)
to_device(self.tq1, train_device)
to_device(self.tq2, train_device)
def prep_rollout(self, rollout_device):
self.pi.eval()
self.q1.eval()
self.q2.eval()
to_device(self.pi, rollout_device)
to_device(self.q1, rollout_device)
to_device(self.q2, rollout_device)
def prep_training(self, train_device):
self.discriminator.train()
to_device(self.discriminator, train_device)
def prep_rollout(self, rollout_device):
self.discriminator.eval()
to_device(self.discriminator, rollout_device)
def prep_training(self, train_device):
self.classifier.train()
to_device(self.classifier, train_device)
def prep_rollout(self, rollout_device):
self.classifier.eval()
to_device(self.classifier, rollout_device)
def prep_training(self, train_device):
self.actor.train()
self.critic.train()
to_device(self.actor, train_device)
to_device(self.critic, train_device)
def prep_rollout(self, rollout_device):
self.actor.eval()
to_device(model=self.actor, device=rollout_device)
def prep_rollout(self, rollout_device):
self.q1.eval()
self.q2.eval()
to_device(self.q1, rollout_device)
to_device(self.q2, rollout_device)
to_device(self.log_alpha, rollout_device)