V2.1 Add spin penalty and reward adjust

Build-ParallelEnv-Target-OffPolicy-SingleStack-SideChannel-EndReward-Easy-V2.1
add spin penalty while agent keep spin will give a penalty reward.
lower Go target in area reward.
This commit is contained in:
2022-12-04 08:40:23 +09:00
parent 0d60c857b7
commit fa90ac3bb6
6 changed files with 185 additions and 134 deletions
+72 -56
View File
@@ -2744,7 +2744,7 @@ MonoBehaviour:
m_Name:
m_EditorClassIdentifier:
m_BrainParameters:
VectorObservationSize: 38
VectorObservationSize: 39
NumStackedVectorObservations: 1
m_ActionSpec:
m_NumContinuousActions: 1
@@ -8228,13 +8228,15 @@ MonoBehaviour:
timeLimit: 30
lockCameraX: 0
lockCameraY: 1
chartOn: 0
chartOn: 1
spinRecordMax: 40
spinPenaltyThreshold: 10
hitTargetRewardDefault: 100
killTargetEnemyRewardDefault: 100
inAreaRewardDefault: 10
inAreaRewardDefault: 2
freeTimeBonusPerSec: 1
targetTimeBonusPerSec: 2
areaTimeBonusPerSec: 1
targetTimeBonusPerSec: 1
areaTimeBonusPerSec: 0.2
distanceReward: 20
goWinRewardDefault: 999
attackWinRewardDefault: 999
@@ -8247,8 +8249,8 @@ MonoBehaviour:
killReward: 5
shootWithoutReadyReward: -0.15
killBonusReward: 0
speedPanalty: 0
viewPanalty: 0
speedPenalty: 0
spinPenalty: 0.04
hitTargetReward: 60
killTargetEnemyReward: 80
inAreaReward: 5
@@ -9550,13 +9552,15 @@ MonoBehaviour:
timeLimit: 30
lockCameraX: 0
lockCameraY: 1
chartOn: 0
chartOn: 1
spinRecordMax: 40
spinPenaltyThreshold: 10
hitTargetRewardDefault: 100
killTargetEnemyRewardDefault: 100
inAreaRewardDefault: 10
inAreaRewardDefault: 2
freeTimeBonusPerSec: 1
targetTimeBonusPerSec: 2
areaTimeBonusPerSec: 1
targetTimeBonusPerSec: 1
areaTimeBonusPerSec: 0.2
distanceReward: 20
goWinRewardDefault: 999
attackWinRewardDefault: 999
@@ -9569,8 +9573,8 @@ MonoBehaviour:
killReward: 5
shootWithoutReadyReward: -0.15
killBonusReward: 0
speedPanalty: 0
viewPanalty: 0
speedPenalty: 0
spinPenalty: 0.04
hitTargetReward: 60
killTargetEnemyReward: 80
inAreaReward: 5
@@ -17322,7 +17326,7 @@ MonoBehaviour:
m_Name:
m_EditorClassIdentifier:
m_BrainParameters:
VectorObservationSize: 38
VectorObservationSize: 39
NumStackedVectorObservations: 1
m_ActionSpec:
m_NumContinuousActions: 1
@@ -19592,13 +19596,15 @@ MonoBehaviour:
timeLimit: 30
lockCameraX: 0
lockCameraY: 1
chartOn: 0
chartOn: 1
spinRecordMax: 40
spinPenaltyThreshold: 10
hitTargetRewardDefault: 100
killTargetEnemyRewardDefault: 100
inAreaRewardDefault: 10
inAreaRewardDefault: 2
freeTimeBonusPerSec: 1
targetTimeBonusPerSec: 2
areaTimeBonusPerSec: 1
targetTimeBonusPerSec: 1
areaTimeBonusPerSec: 0.2
distanceReward: 20
goWinRewardDefault: 999
attackWinRewardDefault: 999
@@ -19611,8 +19617,8 @@ MonoBehaviour:
killReward: 5
shootWithoutReadyReward: -0.15
killBonusReward: 0
speedPanalty: 0
viewPanalty: 0
speedPenalty: 0
spinPenalty: 0.04
hitTargetReward: 60
killTargetEnemyReward: 80
inAreaReward: 5
@@ -25421,7 +25427,7 @@ MonoBehaviour:
m_Name:
m_EditorClassIdentifier:
m_BrainParameters:
VectorObservationSize: 38
VectorObservationSize: 39
NumStackedVectorObservations: 1
m_ActionSpec:
m_NumContinuousActions: 1
@@ -30251,7 +30257,7 @@ MonoBehaviour:
m_Name:
m_EditorClassIdentifier:
m_BrainParameters:
VectorObservationSize: 38
VectorObservationSize: 39
NumStackedVectorObservations: 1
m_ActionSpec:
m_NumContinuousActions: 1
@@ -30866,13 +30872,15 @@ MonoBehaviour:
timeLimit: 30
lockCameraX: 0
lockCameraY: 1
chartOn: 0
chartOn: 1
spinRecordMax: 40
spinPenaltyThreshold: 10
hitTargetRewardDefault: 100
killTargetEnemyRewardDefault: 100
inAreaRewardDefault: 10
inAreaRewardDefault: 2
freeTimeBonusPerSec: 1
targetTimeBonusPerSec: 2
areaTimeBonusPerSec: 1
targetTimeBonusPerSec: 1
areaTimeBonusPerSec: 0.2
distanceReward: 20
goWinRewardDefault: 999
attackWinRewardDefault: 999
@@ -30885,8 +30893,8 @@ MonoBehaviour:
killReward: 5
shootWithoutReadyReward: -0.15
killBonusReward: 0
speedPanalty: 0
viewPanalty: 0
speedPenalty: 0
spinPenalty: 0.04
hitTargetReward: 60
killTargetEnemyReward: 80
inAreaReward: 5
@@ -33261,7 +33269,7 @@ MonoBehaviour:
m_Name:
m_EditorClassIdentifier:
m_BrainParameters:
VectorObservationSize: 38
VectorObservationSize: 39
NumStackedVectorObservations: 1
m_ActionSpec:
m_NumContinuousActions: 1
@@ -34532,7 +34540,7 @@ MonoBehaviour:
m_Name:
m_EditorClassIdentifier:
m_BrainParameters:
VectorObservationSize: 38
VectorObservationSize: 39
NumStackedVectorObservations: 1
m_ActionSpec:
m_NumContinuousActions: 1
@@ -35432,7 +35440,7 @@ MonoBehaviour:
m_Name:
m_EditorClassIdentifier:
m_BrainParameters:
VectorObservationSize: 38
VectorObservationSize: 39
NumStackedVectorObservations: 1
m_ActionSpec:
m_NumContinuousActions: 1
@@ -38454,13 +38462,15 @@ MonoBehaviour:
timeLimit: 30
lockCameraX: 0
lockCameraY: 1
chartOn: 0
chartOn: 1
spinRecordMax: 40
spinPenaltyThreshold: 10
hitTargetRewardDefault: 100
killTargetEnemyRewardDefault: 100
inAreaRewardDefault: 10
inAreaRewardDefault: 2
freeTimeBonusPerSec: 1
targetTimeBonusPerSec: 2
areaTimeBonusPerSec: 1
targetTimeBonusPerSec: 1
areaTimeBonusPerSec: 0.2
distanceReward: 20
goWinRewardDefault: 999
attackWinRewardDefault: 999
@@ -38473,8 +38483,8 @@ MonoBehaviour:
killReward: 5
shootWithoutReadyReward: -0.15
killBonusReward: 0
speedPanalty: 0
viewPanalty: 0
speedPenalty: 0
spinPenalty: 0.04
hitTargetReward: 60
killTargetEnemyReward: 80
inAreaReward: 5
@@ -42051,7 +42061,7 @@ MonoBehaviour:
m_Name:
m_EditorClassIdentifier:
m_BrainParameters:
VectorObservationSize: 38
VectorObservationSize: 39
NumStackedVectorObservations: 1
m_ActionSpec:
m_NumContinuousActions: 1
@@ -42937,13 +42947,15 @@ MonoBehaviour:
timeLimit: 30
lockCameraX: 0
lockCameraY: 1
chartOn: 0
chartOn: 1
spinRecordMax: 40
spinPenaltyThreshold: 10
hitTargetRewardDefault: 100
killTargetEnemyRewardDefault: 100
inAreaRewardDefault: 10
inAreaRewardDefault: 2
freeTimeBonusPerSec: 1
targetTimeBonusPerSec: 2
areaTimeBonusPerSec: 1
targetTimeBonusPerSec: 1
areaTimeBonusPerSec: 0.2
distanceReward: 20
goWinRewardDefault: 999
attackWinRewardDefault: 999
@@ -42956,8 +42968,8 @@ MonoBehaviour:
killReward: 5
shootWithoutReadyReward: -0.15
killBonusReward: 0
speedPanalty: 0
viewPanalty: 0
speedPenalty: 0
spinPenalty: 0.04
hitTargetReward: 60
killTargetEnemyReward: 80
inAreaReward: 5
@@ -51336,13 +51348,15 @@ MonoBehaviour:
timeLimit: 30
lockCameraX: 0
lockCameraY: 1
chartOn: 0
chartOn: 1
spinRecordMax: 40
spinPenaltyThreshold: 10
hitTargetRewardDefault: 100
killTargetEnemyRewardDefault: 100
inAreaRewardDefault: 10
inAreaRewardDefault: 2
freeTimeBonusPerSec: 1
targetTimeBonusPerSec: 2
areaTimeBonusPerSec: 1
targetTimeBonusPerSec: 1
areaTimeBonusPerSec: 0.2
distanceReward: 20
goWinRewardDefault: 999
attackWinRewardDefault: 999
@@ -51355,8 +51369,8 @@ MonoBehaviour:
killReward: 5
shootWithoutReadyReward: -0.15
killBonusReward: 0
speedPanalty: 0
viewPanalty: 0
speedPenalty: 0
spinPenalty: 0.04
hitTargetReward: 60
killTargetEnemyReward: 80
inAreaReward: 5
@@ -53598,13 +53612,15 @@ MonoBehaviour:
timeLimit: 30
lockCameraX: 0
lockCameraY: 1
chartOn: 0
chartOn: 1
spinRecordMax: 40
spinPenaltyThreshold: 10
hitTargetRewardDefault: 100
killTargetEnemyRewardDefault: 100
inAreaRewardDefault: 10
inAreaRewardDefault: 2
freeTimeBonusPerSec: 1
targetTimeBonusPerSec: 2
areaTimeBonusPerSec: 1
targetTimeBonusPerSec: 1
areaTimeBonusPerSec: 0.2
distanceReward: 20
goWinRewardDefault: 999
attackWinRewardDefault: 999
@@ -53617,8 +53633,8 @@ MonoBehaviour:
killReward: 5
shootWithoutReadyReward: -0.15
killBonusReward: 0
speedPanalty: 0
viewPanalty: 0
speedPenalty: 0
spinPenalty: 0.04
hitTargetReward: 60
killTargetEnemyReward: 80
inAreaReward: 5