ROMMEO-AC

This repository is the implementation of the paper, A Regularized Opponent Model with Maximum Entropy Objective.
In the corresponding paper, the experiments were conducted on two environments, Iterated Matrix Games and Differential Games.
However, in this implementation, I only implemented for Differential Games.
I referred to the original authors' codes written in TensorFlow and converted the TensorFlow to PyTorch.

For training and getting results:
python main.py

Results:

Return	Policy

---

References:

The repository of the paper

• This repository provides the codes which are written in TensorFlow.