In goal-conditioned offline reinforcement learning, an agent learns from previously collected data to go to an arbitrary goal. Since the offline data only contains a finite number of trajectories, a main challenge is how to generate more data. Goal-swapping generates additional data by switching trajectory goals but while doing so produces a large number of invalid trajectories. To address this issue, we propose prioritized goal-swapping experience replay (PGSER). PGSER uses a pre-trained Q function to assign higher priority weights to goal swapped transitions that allow reaching the goal. In experiments, PGSER significantly improves over baselines in a wide range of benchmark tasks, including challenging previously unsuccessful dexterous in-hand manipulation tasks.
@article{arxiv.2302.07741,
title = {Prioritized offline Goal-swapping Experience Replay},
author = {Wenyan Yang and Joni Pajarinen and Dinging Cai and Joni Kämäräinen},
journal= {arXiv preprint arXiv:2302.07741},
year = {2023}
}