clipping values and updated zombie

Project/Assets/ML-Agents/Examples/PushBlock/Scripts/ZombiePushBlockDeathEnvController.cs

                     item.Agent.AddReward(score);
                 }
             }
-        }
-        StartCoroutine(GoalScoredSwapGroundMaterial(m_PushBlockSettings.goalScoredMaterial, 0.5f));
+            StartCoroutine(GoalScoredSwapGroundMaterial(m_PushBlockSettings.goalScoredMaterial, 0.5f));
-        ResetScene();
+            ResetScene();
+        }
     }
 
     public void ZombieTouchedBlock()

config/ppo/PushBlock.yaml

       batch_size: 128
       buffer_size: 2048
       learning_rate: 0.0003
-      beta: 0.01
+      beta: 0.001
       epsilon: 0.2
       lambd: 0.95
       num_epoch: 3
-      hidden_units: 512
+      hidden_units: 256
       num_layers: 2
       vis_encode_type: simple
     reward_signals:

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

         for name, head in values.items():
             old_val_tensor = old_values[name]
             returns_tensor = returns[name]
-            #clipped_value_estimate = old_val_tensor + torch.clamp(
-            #   head - old_val_tensor, -1 * epsilon, epsilon
-            #)
+            clipped_value_estimate = old_val_tensor + torch.clamp(
+               head - old_val_tensor, -1 * epsilon, epsilon
+            )
-            #v_opt_b = (returns_tensor - clipped_value_estimate) ** 2
-            #value_loss = ModelUtils.masked_mean(torch.max(v_opt_a, v_opt_b), loss_masks)
-            value_loss = ModelUtils.masked_mean(v_opt_a, loss_masks)
+            v_opt_b = (returns_tensor - clipped_value_estimate) ** 2
+            value_loss = ModelUtils.masked_mean(torch.max(v_opt_a, v_opt_b), loss_masks)
+            #value_loss = ModelUtils.masked_mean(v_opt_a, loss_masks)
             value_losses.append(value_loss)
         value_loss = torch.mean(torch.stack(value_losses))
         return value_loss

Project/Assets/ML-Agents/Examples/PushBlock/Prefabs/ZombiePushBlockCollabArea 1.prefab.meta

+fileFormatVersion: 2
+guid: 0ed44cea8ffd043ca8db76dd48deb7a2
+PrefabImporter:
+  externalObjects: {}
+  userData: 
+  assetBundleName: 
+  assetBundleVariant: 

config/ppo/Zombie.yaml

+behaviors:
+  PushBlock:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 2048
+      learning_rate: 0.0003
+      beta: 0.01
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: constant
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 3
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 20000000 #2000000
+    time_horizon: 64
+    summary_freq: 10000
+    threaded: true