Benchmark Environments for Multitask Learning in Continuous Domains

Abstract

As demand drives systems to generalize to various domains and problems, the study of multitask, transfer and lifelong learning has become an increasingly important pursuit. In discrete domains, performance on the Atari game suite has emerged as the de facto benchmark for assessing multitask learning. However, in continuous domains there is a lack of agreement on standard multitask evaluation environments which makes it difficult to compare different approaches fairly. In this work, we describe a benchmark set of tasks that we have developed in an extendable framework based on OpenAI Gym. We run a simple baseline using Trust Region Policy Optimization and release the framework publicly to be expanded and used for the systematic comparison of multitask, transfer, and lifelong learning in continuous domains.

Accepted at Lifelong Learning: A Reinforcement Learning Approach Workshop @ ICML, Sydney, Australia, 2017

References in corpus (6)

• Benchmarking Deep Reinforcement Learning for Continuous Control
• Actor-Mimic: Deep Multitask and Transfer Reinforcement Learning
• Transfer from Simulation to Real World through Learning Deep Inverse Dynamics Model
• Learning Invariant Feature Spaces to Transfer Skills with Reinforcement Learning
• Q-Prop: Sample-Efficient Policy Gradient with An Off-Policy Critic
• Generalizing Skills with Semi-Supervised Reinforcement Learning

Cited by in corpus (11)

• CausalWorld: A Robotic Manipulation Benchmark for Causal Structure and Transfer Learning
• Knowledge Transfer in Multi-Task Deep Reinforcement Learning for Continuous Control
• Generalization in Transfer Learning
• Multi-task Learning for Continuous Control
• Disentangled Skill Embeddings for Reinforcement Learning
• Augmented World Models Facilitate Zero-Shot Dynamics Generalization From a Single Offline Environment
• Continuous Coordination As a Realistic Scenario for Lifelong Learning
• Diversity-Enriched Option-Critic
• Continual Model-Based Reinforcement Learning with Hypernetworks
• Frosting Weights for Better Continual Training
• Learning Robust Options by Conditional Value at Risk Optimization