Merge branch 'master' into hh/develop/loco-crawler-variable-speed

.circleci/config.yml

             . venv/bin/activate
             mkdir test-reports
             pip freeze > test-reports/pip_versions.txt
-            pytest -n 2 --cov=ml-agents --cov=ml-agents-envs --cov=gym-unity --cov-report html --junitxml=test-reports/junit.xml -p no:warnings
+            pytest --cov=ml-agents --cov=ml-agents-envs --cov=gym-unity --cov-report html --junitxml=test-reports/junit.xml -p no:warnings
 
       - run:
           name: Verify there are no hidden/missing metafiles.
             # Need ruby for search-and-replace
             sudo apt-get update
             sudo apt-get install ruby-full
+
+            # install dotnet and the formatter - see https://docs.microsoft.com/en-us/dotnet/core/install/linux-debian#debian-9-
+            pushd ~
+            wget -O - https://packages.microsoft.com/keys/microsoft.asc | gpg --dearmor > microsoft.asc.gpg
+            sudo mv microsoft.asc.gpg /etc/apt/trusted.gpg.d/
+            wget https://packages.microsoft.com/config/debian/9/prod.list
+            sudo mv prod.list /etc/apt/sources.list.d/microsoft-prod.list
+            sudo chown root:root /etc/apt/trusted.gpg.d/microsoft.asc.gpg
+            sudo chown root:root /etc/apt/sources.list.d/microsoft-prod.list
+            popd
+
+            sudo apt-get install -y apt-transport-https && \
+              sudo apt-get update && \
+              sudo apt-get install -y dotnet-sdk-3.1 && \
+              dotnet tool install -g dotnet-format --version 4.1.131201
+
+            echo "Setting up venv"
             python3 -m venv venv
             . venv/bin/activate
             pip install --upgrade pip
           name: Check Code Style using pre-commit
           command: |
             . venv/bin/activate
+            export PATH="$PATH:~/.dotnet/tools"
             pre-commit run --show-diff-on-failure --all-files
 
   markdown_link_check:

.pre-commit-config.yaml

         name: validate release links
         language: script
         entry: utils/validate_release_links.py
-
+    -   id: dotnet-format
+        name: dotnet-format
+        language: script
+        entry: utils/run_dotnet_format.py
+        types: [c#]
 

DevProject/Assets/ML-Agents/Scripts/Tests/Performance/SensorPerformanceTests.cs

 using NUnit.Framework;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Policies;
 using Unity.MLAgents.Sensors;
 using Unity.MLAgents.Sensors.Reflection;
             {
             }
 
-            public override void Heuristic(float[] actionsOut)
+            public override void Heuristic(in ActionBuffers actionsOut)
             {
             }
         }
                 sensor.AddObservation(new Quaternion(1, 2, 3, 4));
             }
 
-            public override void Heuristic(float[] actionsOut)
+            public override void Heuristic(in ActionBuffers actionsOut)
             {
             }
         }
             [Observable]
             public Quaternion QuaternionField = new Quaternion(1, 2, 3, 4);
 
-            public override void Heuristic(float[] actionsOut)
+            public override void Heuristic(in ActionBuffers actionsOut)
             {
             }
         }
                 get { return m_QuaternionField; }
             }
 
-            public override void Heuristic(float[] actionsOut)
+            public override void Heuristic(in ActionBuffers actionsOut)
             {
             }
         }
         {
             using (Measure.ProfilerMarkers(s_Markers))
             {
-                for(var i=0; i<k_MarkerTestSteps; i++)
+                for (var i = 0; i < k_MarkerTestSteps; i++)
                 {
                     RunAgent<CollectObservationsAgent>(k_NumAgentSteps, 7, ObservableAttributeOptions.Ignore);
                 }

Project/Assets/ML-Agents/Editor/DisableBurstFromMenu.cs

-#if UNITY_CLOUD_BUILD
+#if UNITY_CLOUD_BUILD
 using UnityEditor;
 
 public class DisableBurstFromMenu

Project/Assets/ML-Agents/Examples/3DBall/Scripts/Ball3DAgent.cs

+using System;
+using Unity.MLAgents.Actuators;
+using Random = UnityEngine.Random;
 
 public class Ball3DAgent : Agent
 {
         sensor.AddObservation(m_BallRb.velocity);
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
-        var actionZ = 2f * Mathf.Clamp(vectorAction[0], -1f, 1f);
-        var actionX = 2f * Mathf.Clamp(vectorAction[1], -1f, 1f);
+        var actionZ = 2f * Mathf.Clamp(actionBuffers.ContinuousActions[0], -1f, 1f);
+        var actionX = 2f * Mathf.Clamp(actionBuffers.ContinuousActions[1], -1f, 1f);
 
         if ((gameObject.transform.rotation.z < 0.25f && actionZ > 0f) ||
             (gameObject.transform.rotation.z > -0.25f && actionZ < 0f))
         SetResetParameters();
     }
 
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        actionsOut[0] = -Input.GetAxis("Horizontal");
-        actionsOut[1] = Input.GetAxis("Vertical");
+        var continuousActionsOut = actionsOut.ContinuousActions;
+        continuousActionsOut[0] = -Input.GetAxis("Horizontal");
+        continuousActionsOut[1] = Input.GetAxis("Vertical");
     }
 
     public void SetBall()

Project/Assets/ML-Agents/Examples/3DBall/Scripts/Ball3DHardAgent.cs

 using UnityEngine;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Sensors;
 
 public class Ball3DHardAgent : Agent
         sensor.AddObservation((ball.transform.position - gameObject.transform.position));
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        var actionZ = 2f * Mathf.Clamp(vectorAction[0], -1f, 1f);
-        var actionX = 2f * Mathf.Clamp(vectorAction[1], -1f, 1f);
+        var continuousActions = actionBuffers.ContinuousActions;
+        var actionZ = 2f * Mathf.Clamp(continuousActions[0], -1f, 1f);
+        var actionX = 2f * Mathf.Clamp(continuousActions[1], -1f, 1f);
 
         if ((gameObject.transform.rotation.z < 0.25f && actionZ > 0f) ||
             (gameObject.transform.rotation.z > -0.25f && actionZ < 0f))

Project/Assets/ML-Agents/Examples/Basic/Scripts/BasicController.cs

 using UnityEngine;
 using UnityEngine.SceneManagement;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using UnityEngine.Serialization;
 
 /// <summary>
     /// Controls the movement of the GameObject based on the actions received.
     /// </summary>
     /// <param name="vectorAction"></param>
-    public void ApplyAction(float[] vectorAction)
+    public void ApplyAction(ActionSegment<int> vectorAction)
-        var movement = (int)vectorAction[0];
+        var movement = vectorAction[0];
 
         var direction = 0;
 
         if (Academy.Instance.IsCommunicatorOn)
         {
             // Apply the previous step's actions
-            ApplyAction(m_Agent.GetAction());
+            ApplyAction(m_Agent.GetStoredActionBuffers().DiscreteActions);
             m_Agent?.RequestDecision();
         }
         else
                 // Apply the previous step's actions
-                ApplyAction(m_Agent.GetAction());
+                ApplyAction(m_Agent.GetStoredActionBuffers().DiscreteActions);
 
                 m_TimeSinceDecision = 0f;
                 m_Agent?.RequestDecision();

Project/Assets/ML-Agents/Examples/Basic/TFModels/Basic.nn

 vector_observation����

action_masks����

policy_1/concat_2/concatactionaction_output_shape������?
-action_output_shape




memory_size





version_number




trainer_patch_version




trainer_minor_version




trainer_major_version




is_continuous_control




policy_1/add_2/y




policy_1/Sum/reduction_indices




policy_1/add/y



	
$policy/main_graph_0/hidden_0/BiasAdd
�����?
vector_observation#policy/main_graph_0/hidden_0/kernel


-�
!policy/main_graph_0/hidden_0/bias


�
 policy/main_graph_0/hidden_0/Mul2	�����?
$policy/main_graph_0/hidden_0/BiasAddpolicy_1/dense/MatMul
�����?
 policy/main_graph_0/hidden_0/Mulpolicy/dense/kernel

�
<policy_1/dense/MatMul/patch:0


�
policy_1/strided_slice�



�����?
action_maskspolicy_1/Softmax2�����?
policy_1/dense/MatMulpolicy_1/addd�����?policy_1/Softmaxpolicy_1/add/ypolicy_1/Mulf�����?policy_1/addpolicy_1/strided_slicepolicy_1/Sum������?policy_1/Mulpolicy_1/Sum/reduction_indicespolicy_1/truedivg�����?policy_1/Mulpolicy_1/Sumpolicy_1/add_2d�����?policy_1/truedivpolicy_1/add_2/ypolicy_1/Log_12r�����?
policy_1/add_2policy_1/concat_2/concat2�����?
policy_1/Log_1action2�����?
policy_1/concat_2/concat@@@�A���3�?���3��ƽl>_jX=�P��
-��z�>����0>�����K�>s�>��Y>&�j>�M�����=,�����"����>�P�����>�ɹ<����8��c�==�����:)�y�x�>|\-<���xdO���"����{?=���=[�1>�l>>_����->�8���8��
ž=�m>M�/>����H�=/��Nh���b_�4��>�]�܌<��
>�Ӥ>�E��D>�T=�zO�����A�;Qd���Ͼ�{|��D;>Z�`�/�aV���o�>TJ�=Sa��xİ�5!)>�8�=� 侥Z�='�.�O�<��5n�
A��|�=�����#��jfs>1x=��<)���g����K�>>z>���>���=�< Lb���>�ݘ>X=C=^JѾ��վ*�>�L�>�����`��%T�(Qz��'�>��k�����C�|>0�D�S:3<̋�>�D^=K8�Y^�>kX����f=b�=�׍�J9�=������=t:��
-q�>�h`� �>�J��F��4.��
��[@�;��+>}=��m>�В<xxW��ᔾ+�Q���>��=���>���>E�=��j�>�a�<����Ə<���u���,��FF��VG���
T�m^���&<�+g=Z	^>��u��>tFy>�f�=/{M>����վv��x���n)�=9�A>s�J���@>*�	��:�>��>A��P���q�>qI=�Z���H���l���e>��[>O�þ��:�G����(�)��>���>����G�����>��^=
-�?�ٻ&��B�$>AW5>�٧���?s�=�E�>č�>�-/<��¾�{��#!O��u0>z�<>���>��=+�
�r|�kV�fo�>��=F@Z>�j=V	�>
�6b��U�6='�?sզ��2�;M4��@�C>��?�-
?	�
=x/>��ؾ�g5=c��=�v����c>$Bi>m�����I>ؖ��ou<=�o��cb����ؽ�m�EQ�S�K=�7�=�[��Q�>/*��7��>�j��#|>D�����|��=���W����	�>���=M�6=�W�>�<�>��=����>(���o��>U)�>���>�����-��2�>F>�<ؾ[�7?*��>��E>>1?i��>S�>�# ?�?���>uTJ�W#�>+Y澷&)?�D?��L?���x���d��>��>:'�>�u?�?r_�>�9@?j����>��>���Z
�=��7��Ϟ�@�d�v>�>9O�>��z>H���ᝡ���p=��=J��>��?i(�>8���>%>��%��樾9���n����r>6^->�Х=Ng>�r>�ad>,{��
��>1��<1����!>ev�Z�=+-�����A����<��:>�1f�{J���c=��!�	���]3>�]R�+��<��C>DY0�
-Ty=��)��R�>C�>�[�=޽mU�>?��=�(��:��λ�&.{=��f��3/��kK��\�=W�ʻ1��>�����4�=k�)��������=P�#���,���A�sJ,��ՙ>�kY�>����ܼ~I>%�S���ս�k�վW�g>�f>
-ƙ����l��..R>��ʽ�k�=�=����a�P��IO?��I?�C?��F?t�B?;.�CM?o
(��I?o�F?VGD?h�(����SH?�{H?��>��D?�G?�A?E7F?7���"Ǿk0/?W��
ž�-,?X���
k���!?ߥ߾�����,?Ӿ�⾡�?�T?�.�>��"��N�Y�˾5n0?�?�>*��>%��P�����>.?�l
�b�����?����"	��L-?�H�>x?L"�n��>%~�>�g'��X����^�??��<�ݾ��+?LD!<.V<��,������Q"?,�|����,?G��������?���G��W�-?
+action_output_shape




memory_size





version_number




trainer_patch_version




trainer_minor_version




trainer_major_version




is_continuous_control




policy_1/add/y




policy_1/add_2/y




policy_1/Sum/reduction_indices



	
$policy/main_graph_0/hidden_0/BiasAdd
�����?
vector_observation#policy/main_graph_0/hidden_0/kernel


+�
!policy/main_graph_0/hidden_0/bias


�
 policy/main_graph_0/hidden_0/Mul2	�����?
$policy/main_graph_0/hidden_0/BiasAddpolicy_1/dense/MatMul
�����?
 policy/main_graph_0/hidden_0/Mulpolicy/dense/kernel

�
<policy_1/dense/MatMul/patch:0


�
policy_1/strided_slice�



�����?
action_maskspolicy_1/Softmax2�����?
policy_1/dense/MatMulpolicy_1/addd�����?policy_1/Softmaxpolicy_1/add/ypolicy_1/Mulf�����?policy_1/addpolicy_1/strided_slicepolicy_1/Sum������?policy_1/Mulpolicy_1/Sum/reduction_indicespolicy_1/truedivg�����?policy_1/Mulpolicy_1/Sumpolicy_1/add_2d�����?policy_1/truedivpolicy_1/add_2/ypolicy_1/Log_12r�����?
policy_1/add_2policy_1/concat_2/concat2�����?
policy_1/Log_1action2�����?
policy_1/concat_2/concat@@@�A���3���3�? ��>�����=;���
+�>�e����U>X��=t`��_�>���>*p\>�}�=���=\�?=�f�=a�E�nڙ=E�\�!ü�1�#�n=_ ��NJ�����>Ҍ�<��=��=��'=��Ͼ�&>*��D4�>S}}>_WV>��>�jj>(�f>�>��B>��<
+�&>�,>�����\(��ޤ���a��G?>C�>�e,>��C�2D)>ڦ�=�!>YI>��=Pd���F������܎>����{A>P��=����j�?>�R�n�=�Z����\��>`��)=	M��[�s�'�@��ܦ�ډ��m����
+˾��]��=�;�">]|�>���=��#>	RT�}Q>e�O��|���|>�g�>:�>9ke�3V'>/�1�X�v��>�PDJ��>����B=�_n=bW���>��>,���#�H�
+$_�:�}�NƔ>Nf�>z���am��_�>}��>�D�ƺ`>ш�=;��=1k����>��>d�þ#�
+���>�
+�=����[�=�D>��"�=������>�>%�2�Po��}Ȝ�Ƙ�>�8�>�Ҿ��2>���>��y�
+�>J
��m#����<�V�>
�e��X���p_>�X/��>��=,�b>1�=�*>�
��+����9�>�}ѽ�l>�ϲ=	�R��ܥ.>��ʽۈǽ�GD�,��t��Ο�;
+H/>S41����5�=������)>z���$��>�����a�=��%>@�,�3{�>�ܳ�v{½Wl:>�N=�fƽ}�ľ�V�����w���+�+<�ǹ��7�����=,��œH��a��
?%���=��7?OD?�#s>F�>.k?0�>QƝ>��(����?j�վ�R�><��=�0��Pv�=�Vu���	=ũ�>��<���F7;��ǔ>�>@ӑ=��B�
`�>�x���>�GN>K��=z�X;�+<<
+��=���=<���c�M���2�U?�ر�W�9>1�
?#�H>jP>8�|>��K>�?>:E>W����@ň��{��:��>���=0;�>��
?�
���,���>�>|�e�0��?J�7?i�"?���=�#>>�?�. ?�Ծ�p&��'?�׾�]r?z�?v!�xu?������	�~��=���B����C?�Y?
'?U��>�<�>�d:>�6!?%�о�%�W�T<s�<�9�>�,?#����=��+�X�
�$�c?��2�7!2�_/<?/�e?���>(/9?���>�VK?�L�>{����z�=�&?*��kl�?���?��ݾ�J?1��u����H?5Ƀ�ڽ����?W��>�\.?�AC?���>J�K?��'?9f��s�����>�(7>���>��>�;�Rp?�(.>e:v<�Ҩ�����&º>W�׼	��>���>���&�X�>����I3����=�b$�兢��nC�0�=����l�y�ԼAΠ>d�_>��Һ�
�j�𾚆�>�c��i�)>�Aw>	����¿>���>vft���>�/3>�� =�g;���s�Ř>�*=��->�[�<����K�=��> {���½�K�:5�m����� �^��M�p�>�i>p��=}�{�\�=ut��1>	��) ���;?I��é��@?SG?�:?Bm5?��=?�+;?Uu@?@|��I#��iC?7?�J�:?��<?����->?2
+�>���>rT�iK�>�_�>��������/���] ?A�?�>�O�2�>	�>n������d�c)?{
����^�5?L1��S
	���?�w��q�޾ص?������վ��?s\׾y���n%?�9Ͼ�;��f!?a��>& �>���u\�>���>���8���	���$?�b�>

q>�Ծi}�:����(?�����q�8&?�>���>l�����9�����!?

Project/Assets/ML-Agents/Examples/Bouncer/Scripts/BouncerAgent.cs

 using UnityEngine;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Sensors;
 
 public class BouncerAgent : Agent
         sensor.AddObservation(target.transform.localPosition);
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        for (var i = 0; i < vectorAction.Length; i++)
+        var continuousActions = actionBuffers.ContinuousActions;
+        for (var i = 0; i < continuousActions.Length; i++)
-            vectorAction[i] = Mathf.Clamp(vectorAction[i], -1f, 1f);
+            continuousActions[i] = Mathf.Clamp(continuousActions[i], -1f, 1f);
-        var x = vectorAction[0];
-        var y = ScaleAction(vectorAction[1], 0, 1);
-        var z = vectorAction[2];
+        var x = continuousActions[0];
+        var y = ScaleAction(continuousActions[1], 0, 1);
+        var z = continuousActions[2];
-            vectorAction[0] * vectorAction[0] +
-            vectorAction[1] * vectorAction[1] +
-            vectorAction[2] * vectorAction[2]) / 3f);
+            continuousActions[0] * continuousActions[0] +
+            continuousActions[1] * continuousActions[1] +
+            continuousActions[2] * continuousActions[2]) / 3f);
 
         m_LookDir = new Vector3(x, y, z);
     }
         }
     }
 
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        actionsOut[0] = Input.GetAxis("Horizontal");
-        actionsOut[1] = Input.GetKey(KeyCode.Space) ? 1.0f : 0.0f;
-        actionsOut[2] = Input.GetAxis("Vertical");
+        var continuousActionsOut = actionsOut.ContinuousActions;
+        continuousActionsOut[0] = Input.GetAxis("Horizontal");
+        continuousActionsOut[1] = Input.GetKey(KeyCode.Space) ? 1.0f : 0.0f;
+        continuousActionsOut[2] = Input.GetAxis("Vertical");
     }
 
     void Update()

Project/Assets/ML-Agents/Examples/Crawler/Scripts/CrawlerAgent.cs

 using System;
 using UnityEngine;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgentsExamples;
 using Unity.MLAgents.Sensors;
 using Random = UnityEngine.Random;
         // The dictionary with all the body parts in it are in the jdController
         var bpDict = m_JdController.bodyPartsDict;
 
+        var continuousActions = actionBuffers.ContinuousActions;
-        bpDict[leg0Upper].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], 0);
-        bpDict[leg1Upper].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], 0);
-        bpDict[leg2Upper].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], 0);
-        bpDict[leg3Upper].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], 0);
-        bpDict[leg0Lower].SetJointTargetRotation(vectorAction[++i], 0, 0);
-        bpDict[leg1Lower].SetJointTargetRotation(vectorAction[++i], 0, 0);
-        bpDict[leg2Lower].SetJointTargetRotation(vectorAction[++i], 0, 0);
-        bpDict[leg3Lower].SetJointTargetRotation(vectorAction[++i], 0, 0);
+        bpDict[leg0Upper].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], 0);
+        bpDict[leg1Upper].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], 0);
+        bpDict[leg2Upper].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], 0);
+        bpDict[leg3Upper].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], 0);
+        bpDict[leg0Lower].SetJointTargetRotation(continuousActions[++i], 0, 0);
+        bpDict[leg1Lower].SetJointTargetRotation(continuousActions[++i], 0, 0);
+        bpDict[leg2Lower].SetJointTargetRotation(continuousActions[++i], 0, 0);
+        bpDict[leg3Lower].SetJointTargetRotation(continuousActions[++i], 0, 0);
-        bpDict[leg0Upper].SetJointStrength(vectorAction[++i]);
-        bpDict[leg1Upper].SetJointStrength(vectorAction[++i]);
-        bpDict[leg2Upper].SetJointStrength(vectorAction[++i]);
-        bpDict[leg3Upper].SetJointStrength(vectorAction[++i]);
-        bpDict[leg0Lower].SetJointStrength(vectorAction[++i]);
-        bpDict[leg1Lower].SetJointStrength(vectorAction[++i]);
-        bpDict[leg2Lower].SetJointStrength(vectorAction[++i]);
-        bpDict[leg3Lower].SetJointStrength(vectorAction[++i]);
+        bpDict[leg0Upper].SetJointStrength(continuousActions[++i]);
+        bpDict[leg1Upper].SetJointStrength(continuousActions[++i]);
+        bpDict[leg2Upper].SetJointStrength(continuousActions[++i]);
+        bpDict[leg3Upper].SetJointStrength(continuousActions[++i]);
+        bpDict[leg0Lower].SetJointStrength(continuousActions[++i]);
+        bpDict[leg1Lower].SetJointStrength(continuousActions[++i]);
+        bpDict[leg2Lower].SetJointStrength(continuousActions[++i]);
+        bpDict[leg3Lower].SetJointStrength(continuousActions[++i]);
     }
 
     void FixedUpdate()

Project/Assets/ML-Agents/Examples/FoodCollector/Scripts/FoodCollectorAgent.cs

+using System;
+using Unity.MLAgents.Actuators;
+using Random = UnityEngine.Random;
 
 public class FoodCollectorAgent : Agent
 {
         return new Color32(r, g, b, 255);
     }
 
-    public void MoveAgent(float[] act)
+    public void MoveAgent(ActionSegment<int> act)
     {
         m_Shoot = false;
 
         gameObject.GetComponentInChildren<Renderer>().material = normalMaterial;
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        MoveAgent(vectorAction);
+        MoveAgent(actionBuffers.DiscreteActions);
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        actionsOut[0] = 0f;
-        actionsOut[1] = 0f;
-        actionsOut[2] = 0f;
+        var discreteActionsOut = actionsOut.DiscreteActions;
+        discreteActionsOut[0] = 0;
+        discreteActionsOut[1] = 0;
+        discreteActionsOut[2] = 0;
-            actionsOut[2] = 2f;
+            discreteActionsOut[2] = 2;
-            actionsOut[0] = 1f;
+            discreteActionsOut[0] = 1;
-            actionsOut[2] = 1f;
+            discreteActionsOut[2] = 1;
-            actionsOut[0] = 2f;
+            discreteActionsOut[0] = 2;
-        actionsOut[3] = Input.GetKey(KeyCode.Space) ? 1.0f : 0.0f;
+        discreteActionsOut[3] = Input.GetKey(KeyCode.Space) ? 1 : 0;
     }
 
     public override void OnEpisodeBegin()

Project/Assets/ML-Agents/Examples/FoodCollector/Scripts/FoodCollectorSettings.cs

         // Send stats via SideChannel so that they'll appear in TensorBoard.
         // These values get averaged every summary_frequency steps, so we don't
         // need to send every Update() call.
-        if ((Time.frameCount % 100)== 0)
+        if ((Time.frameCount % 100) == 0)
         {
             m_Recorder.Add("TotalScore", totalScore);
         }

Project/Assets/ML-Agents/Examples/GridWorld/Scripts/GridAgent.cs

 using UnityEngine;
 using System.Linq;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using UnityEngine.Serialization;
 
 public class GridAgent : Agent
         m_ResetParams = Academy.Instance.EnvironmentParameters;
     }
 
-    public override void CollectDiscreteActionMasks(DiscreteActionMasker actionMasker)
+    public override void WriteDiscreteActionMask(IDiscreteActionMask actionMask)
-           // Prevents the agent from picking an action that would make it collide with a wall
+            // Prevents the agent from picking an action that would make it collide with a wall
             var positionX = (int)transform.position.x;
             var positionZ = (int)transform.position.z;
             var maxPosition = (int)m_ResetParams.GetWithDefault("gridSize", 5f) - 1;
-                actionMasker.SetMask(0, new []{ k_Left});
+                actionMask.WriteMask(0, new[] { k_Left });
-                actionMasker.SetMask(0, new []{k_Right});
+                actionMask.WriteMask(0, new[] { k_Right });
-                actionMasker.SetMask(0, new []{k_Down});
+                actionMask.WriteMask(0, new[] { k_Down });
-                actionMasker.SetMask(0, new []{k_Up});
+                actionMask.WriteMask(0, new[] { k_Up });
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        var action = Mathf.FloorToInt(vectorAction[0]);
+        var action = actionBuffers.DiscreteActions[0];
 
         var targetPos = transform.position;
         switch (action)
         }
     }
 
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        actionsOut[0] = k_NoAction;
+        var discreteActionsOut = actionsOut.DiscreteActions;
+        discreteActionsOut[0] = k_NoAction;
-            actionsOut[0] = k_Right;
+            discreteActionsOut[0] = k_Right;
-            actionsOut[0] = k_Up;
+            discreteActionsOut[0] = k_Up;
-            actionsOut[0] = k_Left;
+            discreteActionsOut[0] = k_Left;
-            actionsOut[0] = k_Down;
+            discreteActionsOut[0] = k_Down;
         }
     }
 

Project/Assets/ML-Agents/Examples/Hallway/Scripts/HallwayAgent.cs

 using System.Collections;
 using UnityEngine;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Sensors;
 
 public class HallwayAgent : Agent
         m_GroundRenderer.material = m_GroundMaterial;
     }
 
-    public void MoveAgent(float[] act)
+    public void MoveAgent(ActionSegment<int> act)
-        var action = Mathf.FloorToInt(act[0]);
+        var action = act[0];
         switch (action)
         {
             case 1:
         m_AgentRb.AddForce(dirToGo * m_HallwaySettings.agentRunSpeed, ForceMode.VelocityChange);
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        MoveAgent(vectorAction);
+        MoveAgent(actionBuffers.DiscreteActions);
     }
 
     void OnCollisionEnter(Collision col)
         }
     }
 
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        actionsOut[0] = 0;
+        var discreteActionsOut = actionsOut.DiscreteActions;
+        discreteActionsOut[0] = 0;
-            actionsOut[0] = 3;
+            discreteActionsOut[0] = 3;
-            actionsOut[0] = 1;
+            discreteActionsOut[0] = 1;
-            actionsOut[0] = 4;
+            discreteActionsOut[0] = 4;
-            actionsOut[0] = 2;
+            discreteActionsOut[0] = 2;
         }
     }
 

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

 using System.Collections;
 using UnityEngine;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 
 public class PushAgentBasic : Agent
 {
     /// <summary>
     /// Moves the agent according to the selected action.
     /// </summary>
-    public void MoveAgent(float[] act)
+    public void MoveAgent(ActionSegment<int> act)
-        var action = Mathf.FloorToInt(act[0]);
+        var action = act[0];
 
         switch (action)
         {
     /// <summary>
     /// Called every step of the engine. Here the agent takes an action.
     /// </summary>
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        MoveAgent(vectorAction);
+        MoveAgent(actionBuffers.DiscreteActions);
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        actionsOut[0] = 0;
+        var discreteActionsOut = actionsOut.DiscreteActions;
+        discreteActionsOut[0] = 0;
-            actionsOut[0] = 3;
+            discreteActionsOut[0] = 3;
-            actionsOut[0] = 1;
+            discreteActionsOut[0] = 1;
-            actionsOut[0] = 4;
+            discreteActionsOut[0] = 4;
-            actionsOut[0] = 2;
+            discreteActionsOut[0] = 2;
         }
     }
 

Project/Assets/ML-Agents/Examples/Pyramids/Scripts/PyramidAgent.cs

 using UnityEngine;
 using Random = UnityEngine.Random;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Sensors;
 
 public class PyramidAgent : Agent
         }
     }
 
-    public void MoveAgent(float[] act)
+    public void MoveAgent(ActionSegment<int> act)
-        var action = Mathf.FloorToInt(act[0]);
+        var action = act[0];
         switch (action)
         {
             case 1:
         m_AgentRb.AddForce(dirToGo * 2f, ForceMode.VelocityChange);
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        MoveAgent(vectorAction);
+        MoveAgent(actionBuffers.DiscreteActions);
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        actionsOut[0] = 0;
+        var discreteActionsOut = actionsOut.DiscreteActions;
+        discreteActionsOut[0] = 0;
-            actionsOut[0] = 3;
+            discreteActionsOut[0] = 3;
-            actionsOut[0] = 1;
+            discreteActionsOut[0] = 1;
-            actionsOut[0] = 4;
+            discreteActionsOut[0] = 4;
-            actionsOut[0] = 2;
+            discreteActionsOut[0] = 2;
         }
     }
 

Project/Assets/ML-Agents/Examples/Reacher/Scripts/ReacherAgent.cs

 using UnityEngine;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Sensors;
 
 public class ReacherAgent : Agent
     /// <summary>
     /// The agent's four actions correspond to torques on each of the two joints.
     /// </summary>
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        var torqueX = Mathf.Clamp(vectorAction[0], -1f, 1f) * 150f;
-        var torqueZ = Mathf.Clamp(vectorAction[1], -1f, 1f) * 150f;
+        var torqueX = Mathf.Clamp(actionBuffers.ContinuousActions[0], -1f, 1f) * 150f;
+        var torqueZ = Mathf.Clamp(actionBuffers.ContinuousActions[1], -1f, 1f) * 150f;
-        torqueX = Mathf.Clamp(vectorAction[2], -1f, 1f) * 150f;
-        torqueZ = Mathf.Clamp(vectorAction[3], -1f, 1f) * 150f;
+        torqueX = Mathf.Clamp(actionBuffers.ContinuousActions[2], -1f, 1f) * 150f;
+        torqueZ = Mathf.Clamp(actionBuffers.ContinuousActions[3], -1f, 1f) * 150f;
         m_RbB.AddTorque(new Vector3(torqueX, 0f, torqueZ));
     }
 

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/AdjustTrainingTimescale.cs

 
 using UnityEngine;
 
-namespace  MLAgentsExamples
+namespace MLAgentsExamples
 {
     public class AdjustTrainingTimescale : MonoBehaviour
     {

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/DirectionIndicator.cs

-using UnityEngine;
+using UnityEngine;
-       
+
         public bool updatedByAgent; //should this be updated by the agent? If not, it will use local settings
         public Transform transformToFollow; //ex: hips or body
         public Transform targetToLookAt; //target in the scene the indicator will point to

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/JointDriveController.cs

     [System.Serializable]
     public class BodyPart
     {
-        [Header("Body Part Info")][Space(10)] public ConfigurableJoint joint;
+        [Header("Body Part Info")] [Space(10)] public ConfigurableJoint joint;
-        [Header("Ground & Target Contact")][Space(10)]
+        [Header("Ground & Target Contact")]
+        [Space(10)]
         public GroundContact groundContact;
 
         public TargetContact targetContact;
 
-        [Header("Current Joint Settings")][Space(10)]
+        [Header("Current Joint Settings")]
+        [Space(10)]
         public Vector3 currentEularJointRotation;
 
         [HideInInspector] public float currentStrength;
 
-        [Header("Other Debug Info")][Space(10)]
+        [Header("Other Debug Info")]
+        [Space(10)]
         public Vector3 currentJointForce;
 
         public float currentJointForceSqrMag;
 
     public class JointDriveController : MonoBehaviour
     {
-        [Header("Joint Drive Settings")][Space(10)]
+        [Header("Joint Drive Settings")]
+        [Space(10)]
         public float maxJointSpring;
 
         public float jointDampen;

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/ModelOverrider.cs

     /// </summary>
     public class ModelOverrider : MonoBehaviour
     {
-        HashSet<string> k_SupportedExtensions = new HashSet<string>{"nn", "onnx"};
+        HashSet<string> k_SupportedExtensions = new HashSet<string> { "nn", "onnx" };
         const string k_CommandLineModelOverrideFlag = "--mlagents-override-model";
         const string k_CommandLineModelOverrideDirectoryFlag = "--mlagents-override-model-directory";
         const string k_CommandLineModelOverrideExtensionFlag = "--mlagents-override-model-extension";
 
         int TotalCompletedEpisodes
         {
-            get { return m_PreviousAgentCompletedEpisodes + (m_Agent == null ? 0 : m_Agent.CompletedEpisodes);  }
+            get { return m_PreviousAgentCompletedEpisodes + (m_Agent == null ? 0 : m_Agent.CompletedEpisodes); }
         }
 
         int TotalNumSteps
 
         public bool HasOverrides
         {
-            get { return m_BehaviorNameOverrides.Count > 0 || !string.IsNullOrEmpty(m_BehaviorNameOverrideDirectory);  }
+            get { return m_BehaviorNameOverrides.Count > 0 || !string.IsNullOrEmpty(m_BehaviorNameOverrideDirectory); }
         }
 
         public static string GetOverrideBehaviorName(string originalBehaviorName)
             var args = commandLineArgsOverride ?? Environment.GetCommandLineArgs();
             for (var i = 0; i < args.Length; i++)
             {
-                if (args[i] == k_CommandLineModelOverrideFlag && i < args.Length-2)
+                if (args[i] == k_CommandLineModelOverrideFlag && i < args.Length - 2)
-                else if (args[i] == k_CommandLineModelOverrideDirectoryFlag && i < args.Length-1)
+                else if (args[i] == k_CommandLineModelOverrideDirectoryFlag && i < args.Length - 1)
-                else if (args[i] == k_CommandLineModelOverrideExtensionFlag && i < args.Length-1)
+                else if (args[i] == k_CommandLineModelOverrideExtensionFlag && i < args.Length - 1)
                 {
                     m_OverrideExtension = args[i + 1].Trim().ToLower();
                     var isKnownExtension = k_SupportedExtensions.Contains(m_OverrideExtension);
 #endif
                     }
                 }
-                else if (args[i] == k_CommandLineQuitAfterEpisodesFlag && i < args.Length-1)
+                else if (args[i] == k_CommandLineQuitAfterEpisodesFlag && i < args.Length - 1)
                 {
                     Int32.TryParse(args[i + 1], out maxEpisodes);
                 }
             {
                 assetPath = m_BehaviorNameOverrides[behaviorName];
             }
-            else if(!string.IsNullOrEmpty(m_BehaviorNameOverrideDirectory))
+            else if (!string.IsNullOrEmpty(m_BehaviorNameOverrideDirectory))
             {
                 assetPath = Path.Combine(m_BehaviorNameOverrideDirectory, $"{behaviorName}.{m_OverrideExtension}");
             }
             {
                 model = File.ReadAllBytes(assetPath);
             }
-            catch(IOException)
+            catch (IOException)
             {
                 Debug.Log($"Couldn't load file {assetPath} at full path {Path.GetFullPath(assetPath)}", this);
                 // Cache the null so we don't repeatedly try to load a missing file
 
             if (!overrideOk && m_QuitOnLoadFailure)
             {
-                if(!string.IsNullOrEmpty(overrideError))
+                if (!string.IsNullOrEmpty(overrideError))
                 {
                     Debug.LogWarning(overrideError);
                 }

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/Monitor.cs

-using System.Collections.Generic;
+using System.Collections.Generic;
 using System.Linq;
 using UnityEngine;
 
                 var displayValues = s_DisplayTransformValues[target];
 
                 var index = 0;
-                var orderedKeys = displayValues.Keys.OrderBy(x => - displayValues[x].time);
+                var orderedKeys = displayValues.Keys.OrderBy(x => -displayValues[x].time);
                 foreach (var key in orderedKeys)
                 {
                     s_KeyStyle.alignment = TextAnchor.MiddleRight;
             s_RedStyle = s_ColorStyle[5];
         }
     }
-}
+}

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/OrientationCubeController.cs

-using UnityEngine;
+using UnityEngine;
 
 namespace Unity.MLAgentsExamples
 {

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/ProjectSettingsOverrides.cs

             m_OriginalMaximumDeltaTime = Time.maximumDeltaTime;
             m_OriginalSolverIterations = Physics.defaultSolverIterations;
             m_OriginalSolverVelocityIterations = Physics.defaultSolverVelocityIterations;
-            m_OriginalReuseCollisionCallbacks = Physics.reuseCollisionCallbacks ;
+            m_OriginalReuseCollisionCallbacks = Physics.reuseCollisionCallbacks;
 
             // Override
             Physics.gravity *= gravityMultiplier;

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/SensorBase.cs

         }
 
         /// <inheritdoc/>
-        public void Update() {}
+        public void Update() { }
 
         /// <inheritdoc/>
         public void Reset() { }

Project/Assets/ML-Agents/Examples/SharedAssets/Scripts/TargetController.cs

-using UnityEngine;
+using UnityEngine;
 using Random = UnityEngine.Random;
 using Unity.MLAgents;
 using UnityEngine.Events;
     /// </summary>
     public class TargetController : MonoBehaviour
     {
-        
+
-        
+
-        
+
         [Header("Target Fell Protection")]
         public bool respawnIfFallsOffPlatform = true; //If the target falls off the platform, reset the position.
         public float fallDistance = 5; //distance below the starting height that will trigger a respawn 

Project/Assets/ML-Agents/Examples/Soccer/Scripts/AgentSoccer.cs

 using System;
 using UnityEngine;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Policies;
 
 public class AgentSoccer : Agent
         m_ResetParams = Academy.Instance.EnvironmentParameters;
     }
 
-    public void MoveAgent(float[] act)
+    public void MoveAgent(ActionSegment<int> act)
     {
         var dirToGo = Vector3.zero;
         var rotateDir = Vector3.zero;
-        var forwardAxis = (int)act[0];
-        var rightAxis = (int)act[1];
-        var rotateAxis = (int)act[2];
+        var forwardAxis = act[0];
+        var rightAxis = act[1];
+        var rotateAxis = act[2];
 
         switch (forwardAxis)
         {
             ForceMode.VelocityChange);
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
     {
 
         if (position == Position.Goalie)
             // Existential penalty cumulant for Generic
             timePenalty -= m_Existential;
         }
-        MoveAgent(vectorAction);
+        MoveAgent(actionBuffers.DiscreteActions);
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        Array.Clear(actionsOut, 0, actionsOut.Length);
+        var discreteActionsOut = actionsOut.DiscreteActions;
+        discreteActionsOut.Clear();
-            actionsOut[0] = 1f;
+            discreteActionsOut[0] = 1;
-            actionsOut[0] = 2f;
+            discreteActionsOut[0] = 2;
-            actionsOut[2] = 1f;
+            discreteActionsOut[2] = 1;
-            actionsOut[2] = 2f;
+            discreteActionsOut[2] = 2;
-            actionsOut[1] = 1f;
+            discreteActionsOut[1] = 1;
-            actionsOut[1] = 2f;
+            discreteActionsOut[1] = 2;
         }
     }
     /// <summary>

Project/Assets/ML-Agents/Examples/Startup/Scripts/Startup.cs

             // no scene environment variable is found.
             var args = Environment.GetCommandLineArgs();
             Console.WriteLine("Command line arguments passed: " + String.Join(" ", args));
-            for (int i = 0; i < args.Length; i++) {
-                if (args [i] == k_SceneCommandLineFlag && i < args.Length - 1) {
+            for (int i = 0; i < args.Length; i++)
+            {
+                if (args[i] == k_SceneCommandLineFlag && i < args.Length - 1)
+                {
                     sceneName = args[i + 1];
                 }
             }

Project/Assets/ML-Agents/Examples/Template/Scripts/TemplateAgent.cs

 using UnityEngine;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Sensors;
 
 public class TemplateAgent : Agent
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
     {
     }
 

Project/Assets/ML-Agents/Examples/Tennis/Scripts/HitWall.cs

     public bool net;
 
     public enum FloorHit
-        {
-            Service,
-            FloorHitUnset,
-            FloorAHit,
-            FloorBHit
-        }
+    {
+        Service,
+        FloorHitUnset,
+        FloorAHit,
+        FloorBHit
+    }
 
     public FloorHit lastFloorHit;
 
         lastFloorHit = FloorHit.Service;
         net = false;
     }
-    
+
     void AgentAWins()
     {
         m_AgentA.SetReward(1);

Project/Assets/ML-Agents/Examples/Tennis/Scripts/TennisAgent.cs

 using UnityEngine;
 using UnityEngine.UI;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Sensors;
 
 public class TennisAgent : Agent
         sensor.AddObservation(m_InvertMult * gameObject.transform.rotation.z);
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        var moveX = Mathf.Clamp(vectorAction[0], -1f, 1f) * m_InvertMult;
-        var moveY = Mathf.Clamp(vectorAction[1], -1f, 1f);
-        var rotate = Mathf.Clamp(vectorAction[2], -1f, 1f) * m_InvertMult;
+        var continuousActions = actionBuffers.ContinuousActions;
+        var moveX = Mathf.Clamp(continuousActions[0], -1f, 1f) * m_InvertMult;
+        var moveY = Mathf.Clamp(continuousActions[1], -1f, 1f);
+        var rotate = Mathf.Clamp(continuousActions[2], -1f, 1f) * m_InvertMult;
 
         if (moveY > 0.5 && transform.position.y - transform.parent.transform.position.y < -1.5f)
         {
         m_TextComponent.text = score.ToString();
     }
 
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        actionsOut[0] = Input.GetAxis("Horizontal");    // Racket Movement
-        actionsOut[1] = Input.GetKey(KeyCode.Space) ? 1f : 0f;   // Racket Jumping
-        actionsOut[2] = Input.GetAxis("Vertical");   // Racket Rotation
+        var continuousActionsOut = actionsOut.ContinuousActions;
+        continuousActionsOut[0] = Input.GetAxis("Horizontal");    // Racket Movement
+        continuousActionsOut[1] = Input.GetKey(KeyCode.Space) ? 1f : 0f;   // Racket Jumping
+        continuousActionsOut[2] = Input.GetAxis("Vertical");   // Racket Rotation
     }
 
     public override void OnEpisodeBegin()

Project/Assets/ML-Agents/Examples/Walker/Scripts/WalkerAgent.cs

 using System;
 using UnityEngine;
 using Unity.MLAgents;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgentsExamples;
 using Unity.MLAgents.Sensors;
 using BodyPart = Unity.MLAgentsExamples.BodyPart;
         }
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        bpDict[chest].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], vectorAction[++i]);
-        bpDict[spine].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], vectorAction[++i]);
+        var continuousActions = actionBuffers.ContinuousActions;
+        bpDict[chest].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], continuousActions[++i]);
+        bpDict[spine].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], continuousActions[++i]);
-        bpDict[thighL].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], 0);
-        bpDict[thighR].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], 0);
-        bpDict[shinL].SetJointTargetRotation(vectorAction[++i], 0, 0);
-        bpDict[shinR].SetJointTargetRotation(vectorAction[++i], 0, 0);
-        bpDict[footR].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], vectorAction[++i]);
-        bpDict[footL].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], vectorAction[++i]);
+        bpDict[thighL].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], 0);
+        bpDict[thighR].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], 0);
+        bpDict[shinL].SetJointTargetRotation(continuousActions[++i], 0, 0);
+        bpDict[shinR].SetJointTargetRotation(continuousActions[++i], 0, 0);
+        bpDict[footR].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], continuousActions[++i]);
+        bpDict[footL].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], continuousActions[++i]);
-        bpDict[armL].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], 0);
-        bpDict[armR].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], 0);
-        bpDict[forearmL].SetJointTargetRotation(vectorAction[++i], 0, 0);
-        bpDict[forearmR].SetJointTargetRotation(vectorAction[++i], 0, 0);
-        bpDict[head].SetJointTargetRotation(vectorAction[++i], vectorAction[++i], 0);
+        bpDict[armL].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], 0);
+        bpDict[armR].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], 0);
+        bpDict[forearmL].SetJointTargetRotation(continuousActions[++i], 0, 0);
+        bpDict[forearmR].SetJointTargetRotation(continuousActions[++i], 0, 0);
+        bpDict[head].SetJointTargetRotation(continuousActions[++i], continuousActions[++i], 0);
-        bpDict[chest].SetJointStrength(vectorAction[++i]);
-        bpDict[spine].SetJointStrength(vectorAction[++i]);
-        bpDict[head].SetJointStrength(vectorAction[++i]);
-        bpDict[thighL].SetJointStrength(vectorAction[++i]);
-        bpDict[shinL].SetJointStrength(vectorAction[++i]);
-        bpDict[footL].SetJointStrength(vectorAction[++i]);
-        bpDict[thighR].SetJointStrength(vectorAction[++i]);
-        bpDict[shinR].SetJointStrength(vectorAction[++i]);
-        bpDict[footR].SetJointStrength(vectorAction[++i]);
-        bpDict[armL].SetJointStrength(vectorAction[++i]);
-        bpDict[forearmL].SetJointStrength(vectorAction[++i]);
-        bpDict[armR].SetJointStrength(vectorAction[++i]);
-        bpDict[forearmR].SetJointStrength(vectorAction[++i]);
+        bpDict[chest].SetJointStrength(continuousActions[++i]);
+        bpDict[spine].SetJointStrength(continuousActions[++i]);
+        bpDict[head].SetJointStrength(continuousActions[++i]);
+        bpDict[thighL].SetJointStrength(continuousActions[++i]);
+        bpDict[shinL].SetJointStrength(continuousActions[++i]);
+        bpDict[footL].SetJointStrength(continuousActions[++i]);
+        bpDict[thighR].SetJointStrength(continuousActions[++i]);
+        bpDict[shinR].SetJointStrength(continuousActions[++i]);
+        bpDict[footR].SetJointStrength(continuousActions[++i]);
+        bpDict[armL].SetJointStrength(continuousActions[++i]);
+        bpDict[forearmL].SetJointStrength(continuousActions[++i]);
+        bpDict[armR].SetJointStrength(continuousActions[++i]);
+        bpDict[forearmR].SetJointStrength(continuousActions[++i]);
     }
 
     //Update OrientationCube and DirectionIndicator

Project/Assets/ML-Agents/Examples/WallJump/Scripts/WallJumpAgent.cs

 using UnityEngine;
 using Unity.MLAgents;
 using Unity.Barracuda;
+using Unity.MLAgents.Actuators;
 using Unity.MLAgents.Sensors;
 using Unity.MLAgentsExamples;
 
         m_GroundRenderer.material = m_GroundMaterial;
     }
 
-    public void MoveAgent(float[] act)
+    public void MoveAgent(ActionSegment<int> act)
     {
         AddReward(-0.0005f);
         var smallGrounded = DoGroundCheck(true);
         var rotateDir = Vector3.zero;
-        var dirToGoForwardAction = (int)act[0];
-        var rotateDirAction = (int)act[1];
-        var dirToGoSideAction = (int)act[2];
-        var jumpAction = (int)act[3];
+        var dirToGoForwardAction = act[0];
+        var rotateDirAction = act[1];
+        var dirToGoSideAction = act[2];
+        var jumpAction = act[3];
 
         if (dirToGoForwardAction == 1)
             dirToGo = (largeGrounded ? 1f : 0.5f) * 1f * transform.forward;
         jumpingTime -= Time.fixedDeltaTime;
     }
 
-    public override void OnActionReceived(float[] vectorAction)
+    public override void OnActionReceived(ActionBuffers actionBuffers)
+
-        MoveAgent(vectorAction);
+        MoveAgent(actionBuffers.DiscreteActions);
         if ((!Physics.Raycast(m_AgentRb.position, Vector3.down, 20))
             || (!Physics.Raycast(m_ShortBlockRb.position, Vector3.down, 20)))
         {
         }
     }
 
-    public override void Heuristic(float[] actionsOut)
+    public override void Heuristic(in ActionBuffers actionsOut)
-        System.Array.Clear(actionsOut, 0, actionsOut.Length);
+        var discreteActionsOut = actionsOut.DiscreteActions;
+        discreteActionsOut.Clear();
-            actionsOut[1] = 2f;
+            discreteActionsOut[1] = 2;
-            actionsOut[0] = 1f;
+            discreteActionsOut[0] = 1;
-            actionsOut[1] = 1f;
+            discreteActionsOut[1] = 1;
-            actionsOut[0] = 2f;
+            discreteActionsOut[0] = 2;
-        actionsOut[3] = Input.GetKey(KeyCode.Space) ? 1.0f : 0.0f;
+        discreteActionsOut[3] = Input.GetKey(KeyCode.Space) ? 1 : 0;
     }
 
     // Detect when the agent hits the goal

Project/Assets/ML-Agents/Examples/Worm/Prefabs/PlatformWormDynamicTarget.prefab

   m_Interpolate: 0
   m_Constraints: 0
   m_CollisionDetection: 0
+--- !u!114 &8042564747579887
+MonoBehaviour:
+  m_ObjectHideFlags: 0
+  m_CorrespondingSourceObject: {fileID: 0}
+  m_PrefabInstance: {fileID: 0}
+  m_PrefabAsset: {fileID: 0}
+  m_GameObject: {fileID: 7516457449653310668}
+  m_Enabled: 1
+  m_EditorHideFlags: 0
+  m_Script: {fileID: 11500000, guid: 3c8f113a8b8d94967b1b1782c549be81, type: 3}
+  m_Name: 
+  m_EditorClassIdentifier: 
+  tagToDetect: agent
+  spawnRadius: 40
+  respawnIfTouched: 1
+  respawnIfFallsOffPlatform: 1