Supplementary Material: Supported Policy Optimization for Offline Reinforcement Learning