def update(self): '''Update the agent after training''' for net in [self.net]: net.update_lr() explore_vars = [self.action_policy_update(self, body) for body in self.agent.nanflat_body_a] explore_var_a = self.nanflat_to_data_a('explore_var', explore_vars) return explore_var_a
def update(self): '''Update the agent after training''' for net_name in self.net_names: net = getattr(self, net_name) net.update_lr(self.body.env.clock) explore_var = self.action_policy_update(self, self.body) return explore_var
def update(self): nets = [self.net] if self.share_architecture else [self.net, self.critic] for net in nets: net.update_lr() explore_vars = [self.action_policy_update(self, body) for body in self.agent.nanflat_body_a] explore_var_a = self.nanflat_to_data_a('explore_var', explore_vars) return explore_var_a
def update(self): '''Update the agent after training''' space_clock = util.s_get(self, 'aeb_space.clock') for net in [self.net]: net.update_lr(space_clock) explore_vars = [self.action_policy_update(self, body) for body in self.agent.nanflat_body_a] explore_var_a = self.nanflat_to_data_a('explore_var', explore_vars) return explore_var_a
def update(self): for net in [self.net]: net.update_lr() explore_vars = [ self.action_policy_update(self, body) for body in self.agent.nanflat_body_a ] explore_var_a = self.nanflat_to_data_a('explore_var', explore_vars) return explore_var_a
def update(self): space_clock = util.s_get(self, 'aeb_space.clock') nets = [self.net ] if self.share_architecture else [self.net, self.critic] for net in nets: net.update_lr(space_clock) explore_vars = [ self.action_policy_update(self, body) for body in self.agent.nanflat_body_a ] explore_var_a = self.nanflat_to_data_a('explore_var', explore_vars) return explore_var_a
def update(self): for net_name in self.net_names: net = getattr(self, net_name) net.update_lr(self.body.env.clock) explore_var = self.action_policy_update(self, self.body) return explore_var