浏览代码

remove hh rew. add trained no-hh model

/active-variablespeed
HH 4 年前
当前提交
8bbfb46c
共有 6 个文件被更改,包括 2027 次插入2 次删除
  1. 2
      Project/Assets/ML-Agents/Examples/Walker/Prefabs/DynamicPlatformWalker.prefab
  2. 3
      Project/Assets/ML-Agents/Examples/Walker/Scripts/WalkerAgent.cs
  3. 1001
      Project/Assets/ML-Agents/Examples/Walker/TFModels/WalkerDyMatchSpeedAvgVelALLRBsRandSpeed_5kDampen_HeavierArms_NO_HHRew.nn
  4. 11
      Project/Assets/ML-Agents/Examples/Walker/TFModels/WalkerDyMatchSpeedAvgVelALLRBsRandSpeed_5kDampen_HeavierArms_NO_HHRew.nn.meta
  5. 1001
      Project/Assets/ML-Agents/Examples/Walker/TFModels/WalkerDynamic CL RndSpeedObserveInvLerpAndWalkSpeedRatio.nn
  6. 11
      Project/Assets/ML-Agents/Examples/Walker/TFModels/WalkerDynamic CL RndSpeedObserveInvLerpAndWalkSpeedRatio.nn.meta

2
Project/Assets/ML-Agents/Examples/Walker/Prefabs/DynamicPlatformWalker.prefab


type: 3}
propertyPath: m_Model
value:
objectReference: {fileID: 11400000, guid: 70a698d5b13784437b971995a3ef994c,
objectReference: {fileID: 11400000, guid: 976f2b5823b0f4ffd933de703cb7e826,
type: 3}
- target: {fileID: 895268871377934297, guid: 765582efd9dda46ed98564603316353f,
type: 3}

3
Project/Assets/ML-Agents/Examples/Walker/Scripts/WalkerAgent.cs


// avgVelValue = velSum/4;
// velInverseLerpVal = VelocityInverseLerp(cubeForward * walkingSpeed, avgVelValue);
velInverseLerpVal = VelocityInverseLerp(cubeForward * walkingSpeed);
rewardManager.UpdateReward("productOfAllRewards", velInverseLerpVal * lookAtTargetReward * headHeightOverFeetReward);
rewardManager.UpdateReward("productOfAllRewards", velInverseLerpVal * lookAtTargetReward);
// rewardManager.UpdateReward("productOfAllRewards", velInverseLerpVal * lookAtTargetReward * headHeightOverFeetReward);
// velInverseLerpVal = VelocityInverseLerp(Vector3.zero, cubeForward * walkingSpeed, avgVelValue);
//This reward will approach 1 if it matches and approach zero as it deviates

1001
Project/Assets/ML-Agents/Examples/Walker/TFModels/WalkerDyMatchSpeedAvgVelALLRBsRandSpeed_5kDampen_HeavierArms_NO_HHRew.nn
文件差异内容过多而无法显示
查看文件

11
Project/Assets/ML-Agents/Examples/Walker/TFModels/WalkerDyMatchSpeedAvgVelALLRBsRandSpeed_5kDampen_HeavierArms_NO_HHRew.nn.meta


fileFormatVersion: 2
guid: 976f2b5823b0f4ffd933de703cb7e826
ScriptedImporter:
fileIDToRecycleName:
11400000: main obj
11400002: model data
externalObjects: {}
userData:
assetBundleName:
assetBundleVariant:
script: {fileID: 11500000, guid: 19ed1486aa27d4903b34839f37b8f69f, type: 3}

1001
Project/Assets/ML-Agents/Examples/Walker/TFModels/WalkerDynamic CL RndSpeedObserveInvLerpAndWalkSpeedRatio.nn
文件差异内容过多而无法显示
查看文件

11
Project/Assets/ML-Agents/Examples/Walker/TFModels/WalkerDynamic CL RndSpeedObserveInvLerpAndWalkSpeedRatio.nn.meta


fileFormatVersion: 2
guid: 5ae39043f159147e6a07d4e59027512a
ScriptedImporter:
fileIDToRecycleName:
11400000: main obj
11400002: model data
externalObjects: {}
userData:
assetBundleName:
assetBundleVariant:
script: {fileID: 11500000, guid: 19ed1486aa27d4903b34839f37b8f69f, type: 3}
正在加载...
取消
保存