Merge remote-tracking branch 'upstream/develop' into develop-flat-code-restructure

MLAgentsSDK/Assets/ML-Agents/Examples/BananaCollectors/Prefabs/VisualRLArea.prefab

   agentParameters:
     agentCameras:
     - {fileID: 20214279307138674}
-    maxStep: 0
+    maxStep: 5000
-    numberOfActionsBetweenDecisions: 4
+    numberOfActionsBetweenDecisions: 5
   myAcademyObj: {fileID: 0}
   area: {fileID: 1575621625717122}
   turnSpeed: 300
   agentParameters:
     agentCameras:
     - {fileID: 20216502261607616}
-    maxStep: 0
+    maxStep: 5000
-    numberOfActionsBetweenDecisions: 4
+    numberOfActionsBetweenDecisions: 5
   myAcademyObj: {fileID: 0}
   area: {fileID: 1575621625717122}
   turnSpeed: 300
   agentParameters:
     agentCameras:
     - {fileID: 20704335789641980}
-    maxStep: 0
+    maxStep: 5000
-    numberOfActionsBetweenDecisions: 4
+    numberOfActionsBetweenDecisions: 5
   myAcademyObj: {fileID: 0}
   area: {fileID: 1575621625717122}
   turnSpeed: 300
   agentParameters:
     agentCameras:
     - {fileID: 20130538700979288}
-    maxStep: 0
+    maxStep: 5000
-    numberOfActionsBetweenDecisions: 4
+    numberOfActionsBetweenDecisions: 5
   myAcademyObj: {fileID: 0}
   area: {fileID: 1575621625717122}
   turnSpeed: 300

MLAgentsSDK/Assets/ML-Agents/Examples/BananaCollectors/Scenes/VisualBanana.unity

       propertyPath: m_RootOrder
       value: 4
       objectReference: {fileID: 0}
-    - target: {fileID: 114813974235827884, guid: 79d09bf6cbfa64f55a572755f594652f,
+    - target: {fileID: 114395305755641442, guid: 79d09bf6cbfa64f55a572755f594652f,
-    - target: {fileID: 114813974235827884, guid: 79d09bf6cbfa64f55a572755f594652f,
+    - target: {fileID: 114395305755641442, guid: 79d09bf6cbfa64f55a572755f594652f,
-    - target: {fileID: 114661537450162482, guid: 79d09bf6cbfa64f55a572755f594652f,
+    - target: {fileID: 114863868205103286, guid: 79d09bf6cbfa64f55a572755f594652f,
-    - target: {fileID: 114661537450162482, guid: 79d09bf6cbfa64f55a572755f594652f,
+    - target: {fileID: 114863868205103286, guid: 79d09bf6cbfa64f55a572755f594652f,
-    - target: {fileID: 114863868205103286, guid: 79d09bf6cbfa64f55a572755f594652f,
+    - target: {fileID: 114661537450162482, guid: 79d09bf6cbfa64f55a572755f594652f,
-    - target: {fileID: 114863868205103286, guid: 79d09bf6cbfa64f55a572755f594652f,
+    - target: {fileID: 114661537450162482, guid: 79d09bf6cbfa64f55a572755f594652f,
-    - target: {fileID: 114395305755641442, guid: 79d09bf6cbfa64f55a572755f594652f,
+    - target: {fileID: 114813974235827884, guid: 79d09bf6cbfa64f55a572755f594652f,
-    - target: {fileID: 114395305755641442, guid: 79d09bf6cbfa64f55a572755f594652f,
+    - target: {fileID: 114813974235827884, guid: 79d09bf6cbfa64f55a572755f594652f,
         type: 2}
       propertyPath: myAcademyObj
       value: 

MLAgentsSDK/Assets/ML-Agents/Scripts/ActionMasker.cs

         /// actions.</returns>
         public bool[] GetMask()
         {
-            AssertMask();
+            if (_currentMask != null)
+            {
+                AssertMask();
+            }
             return _currentMask;
         }
 

trainer_config.yaml

     batch_size: 1024
     beta: 5.0e-3
     buffer_size: 10240
+    max_steps: 1.0e5
 
 BouncerBrain:
     normalize: true
 
 VisualPyramidBrain:
     use_curiosity: true
+    curiosity_strength: 0.01
+    curiosity_enc_size: 256
-    batch_size: 32
-    buffer_size: 1024
+    batch_size: 64
+    buffer_size: 2024
-    num_layers: 2
+    num_layers: 1
     beta: 1.0e-2
     max_steps: 5.0e5
     num_epoch: 3
     num_epoch: 3
     buffer_size: 1024
     batch_size: 128
+    max_steps: 5.0e5
+    summary_freq: 1000
+    time_horizon: 64
+
+VisualHallwayBrain:
+    use_recurrent: true
+    sequence_length: 64
+    num_layers: 1
+    hidden_units: 128
+    memory_size: 256
+    beta: 1.0e-2
+    gamma: 0.99
+    num_epoch: 3
+    buffer_size: 1024
+    batch_size: 64
+    max_steps: 5.0e5
+    summary_freq: 1000
+    time_horizon: 64
+
+VisualPushBlockBrain:
+    use_recurrent: true
+    sequence_length: 32
+    num_layers: 1
+    hidden_units: 128
+    memory_size: 256
+    beta: 1.0e-2
+    gamma: 0.99
+    num_epoch: 3
+    buffer_size: 1024
+    batch_size: 64
     max_steps: 5.0e5
     summary_freq: 1000
     time_horizon: 64