_

ml-agents/mlagents/trainers/sac/optimizer_torch.py

         policy_loss = self.sac_policy_loss(log_probs, q1p_out, masks)
         entropy_loss = self.sac_entropy_loss(log_probs, masks)
 
-
-
-
         # Compute surrogate loss for predicting cube position :
 
         l_1 = self.value_network.q1_network.network_body.get_surrogate_loss(current_obs)
 
-        surrogate_loss_p = self.policy.actor_critic.network_body.get_surrogate_loss(current_obs) * 0.05
+        surrogate_loss_p = (
+            self.policy.actor_critic.network_body.get_surrogate_loss(current_obs) * 0.05
+        )
-
-
-
 
         total_value_loss = q1_loss + q2_loss + value_loss
 

ml-agents/mlagents/trainers/torch/networks.py

             normalize=self.normalize,
         )
 
-
-
 
         self.linear_encoder = LinearEncoder(
             total_enc_size, network_settings.num_layers, self.h_size
         actions: Optional[torch.Tensor] = None,
         memories: Optional[torch.Tensor] = None,
         sequence_length: int = 1,
-        retrun_target = False
+        retrun_target=False,
     ) -> Tuple[torch.Tensor, torch.Tensor]:
         encodes = []
         for idx, processor in enumerate(self.processors):
         loss = torch.sum((prediction - target) ** 2, dim=1)
         loss = torch.mean(loss)
         return loss
-
 
 
 class ValueNetwork(nn.Module):