What Matters for Adversarial Imitation Learning?
Manu Orsini, Anton Raichuk, Leonard Hussenot, Damien Vincent, Robert Dadashi, Sertan Girgin, Matthieu Geist, Olivier Bachem
Abstract
Adversarial imitation learning has become a popular framework for imitation in continuous control. Over the years, several variations of its components were proposed to enhance the performance of the learned policies as well as the sample complexity of the algorithm. In practice, these choices are rarely tested all together in rigorous empirical studies. It is therefore difficult to discuss and understand what choices, among the high-level algorithmic options as well as low-level implementation details, matter. To tackle this issue, we implement more than 50 of these choices in a generic adversarial imitation learning framework and investigate their impacts in a large-scale study (>500k trained agents) with both synthetic and human-generated demonstrations. We analyze the key results and highlight the most surprising findings.
BibTeX
@inproceedings{
orsini2021what,
title={What Matters for Adversarial Imitation Learning?},
author={Manu Orsini and Anton Raichuk and Leonard Hussenot and Damien Vincent and Robert Dadashi and Sertan Girgin and Matthieu Geist and Olivier Bachem and Olivier Pietquin and Marcin Andrychowicz},
booktitle={Advances in Neural Information Processing Systems},
editor={A. Beygelzimer and Y. Dauphin and P. Liang and J. Wortman Vaughan},
year={2021},
url={https://openreview.net/forum?id=-OrwaD3bG91}
}