Continuous Coordination As a Realistic Scenario for Lifelong Learning

By Hadi Nekoei, Akilesh Badrinaaraayanan, Aaron C. Courville, Sarath Chandar. 2021

52 citations recorded by the source corpus.

Current deep reinforcement learning (RL) algorithms are still highly task-specific and lack the ability to generalize to new environments. Lifelong learning (LLL), however, aims at solving multiple tasks sequentially by efficiently transferring and using knowledge between tasks. Despite a surge of interest in lifelong RL in recent years, the lack of a realistic testbed makes robust evaluation of LLL algorithms difficult. Multi-agent RL (MARL), on the other hand, can be seen as a natural scenario for lifelong RL due to its inherent non-stationarity, since the agents' policies change over time. In this work, we introduce a multi-agent lifelong learning testbed that supports both zero-shot and few-shot settings. Our setup is based on Hanabi -- a partially-observable, fully cooperative multi-agent game that has been shown to be challenging for zero-shot coordination. Its large strategy space makes it a desirable environment for lifelong RL tasks. We evaluate several recent MARL methods, and benchmark state-of-the-art LLL algorithms in limited memory and computation regimes to shed light on their strengths and weaknesses. This continual learning paradigm also provides us with a pragmatic way of going beyond centralized training which is the most commonly used training protocol in MARL. We empirically show that the agents trained in our setup are able to coordinate well with unseen agents, without any additional assumptions made by previous works. The code and all pre-trained models are available at https://github.com/chandar-lab/Lifelong-Hanabi.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Meta & Continual Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Theory and Mathematics

In Meta & Continual Learning

Citations in the corpus

A Comprehensive Survey of Continual Learning: Theory, Method and Application (2023, background): “A distinctive feature of CRL is its high diversity of applications and benchmarks, including continuous control [203], [279], [296], [449], maze navigation [274], [370], video games (Atari [10], [59], [195], [220], [370], [429], StarCraft [372], Minecraft [405], Hanabi [312]), etc.”
Learning with Recoverable Forgetting (2022, background): “Apart from those works that still need part of the old data, many researchers are devoted to developing the methods without storing the old data by synthesizing old data [35, 39] or even without referring to any old data [37, 36, 27].”
Same State, Different Task: Continual Reinforcement Learning without Interference (2021, methodology): “To alleviate forgetting we consider simple weight space [35] and functional regularizations [29] as these have been repeatedly shown to be effective in CRL [45].”
Continual World: A Robotic Benchmark For Continual Reinforcement Learning (2021, background): “Lifelong Hanabi [36] consider a multi-agent reinforcement learning setting based on Hanabi, a cooperative game requiring significant coordination between agents.”
CORA: Benchmarks, Baselines, and Metrics as a Platform for Continual Reinforcement Learning Agents (2021, background): “Nekoei et al. (2021) present Lifelong Hanabi for continual learning in a multi-agent RL environemnt.”
Model-Free Generative Replay for Lifelong Reinforcement Learning: Application to Starcraft-2 (2022, methodology): “Other interesting avenues include the application of the proposed system and techniques to multi-agent RL (Nekoei et al., 2021) and federated learning (Rostami et al.”
How to Reuse and Compose Knowledge for a Lifetime of Tasks: A Survey on Continual Learning and Functional Composition (2022, background): “In the continual learning setting, a recent benchmark evaluates approaches in a multiagent coordination setting (Nekoei et al., 2021).”