Using Offline Data to Speed Up Reinforcement Learning in Procedurally Generated Environments

TL;DR

This paper investigates how offline data, through imitation learning, can enhance sample efficiency in reinforcement learning within procedurally generated environments, demonstrating significant improvements with minimal offline trajectories.

Contribution

The study shows that offline imitation learning, used for pre-training or concurrently with online RL, significantly boosts sample efficiency and can achieve optimal policies with very few offline trajectories.

Findings

Offline IL improves sample efficiency in RL tasks.

Pre-training with just two trajectories can enable learning of optimal policies.

Concurrent IL during RL training consistently enhances convergence to optimal policies.

Abstract

One of the key challenges of Reinforcement Learning (RL) is the ability of agents to generalise their learned policy to unseen settings. Moreover, training RL agents requires large numbers of interactions with the environment. Motivated by the recent success of Offline RL and Imitation Learning (IL), we conduct a study to investigate whether agents can leverage offline data in the form of trajectories to improve the sample-efficiency in procedurally generated environments. We consider two settings of using IL from offline data for RL: (1) pre-training a policy before online RL training and (2) concurrently training a policy with online RL and IL from offline data. We analyse the impact of the quality (optimality of trajectories) and diversity (number of trajectories and covered level) of available offline trajectories on the effectiveness of both approaches. Across four well-known sparse reward tasks in the MiniGrid environment, we find that using IL for pre-training and concurrently during online RL training both consistently improve the sample-efficiency while converging to optimal policies. Furthermore, we show that pre-training a policy from as few as two trajectories can make the difference between learning an optimal policy at the end of online training and not learning at all. Our findings motivate the widespread adoption of IL for pre-training and concurrent IL in procedurally generated environments whenever offline trajectories are available or can be generated.