List your company

Continuous control locomotion on MuJoCo Humanoid v3 (train)

665Avg Performance (1M Steps)GePPO
580.76602.63624.5646.37Oct 29, 2021

Evaluation results

MethodLinks
GePPOclipping parameter=0.1, value function estimator=V-trace2021.106651,3450.8514
PPOclipping parameter=0.2, batch size=20482021.105849721—