Pytorch implementation for " weighted policy constraints for offline reinforcement learning".
This implementation is build on the official TD3+BC code, and only add several new lines to TD3+BC code to get gain a siganificant performance improvements.
python main.py