Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning

By Tianhe Yu, Deirdre Quillen, Zhanpeng He, Ryan C. Julian, Karol Hausman, Chelsea Finn, Sergey Levine. 2019

1,715 citations recorded by the source corpus.

Meta-reinforcement learning algorithms can enable robots to acquire new skills much more quickly, by leveraging prior experience to learn how to learn. However, much of the current research on meta-reinforcement learning focuses on task distributions that are very narrow. For example, a commonly used meta-reinforcement learning benchmark uses different running velocities for a simulated robot as different tasks. When policies are meta-trained on such narrow task distributions, they cannot possibly generalize to more quickly acquire entirely new tasks. Therefore, if the aim of these methods is to enable faster acquisition of entirely new behaviors, we must evaluate them on task distributions that are sufficiently broad to enable generalization to new behaviors. In this paper, we propose an open-source simulated benchmark for meta-reinforcement learning and multi-task learning consisting of 50 distinct robotic manipulation tasks. Our aim is to make it possible to develop algorithms that generalize to accelerate the acquisition of entirely new, held-out tasks. We evaluate 6 state-of-the-art meta-reinforcement learning and multi-task learning algorithms on these tasks. Surprisingly, while each task and its variations (e.g., with different object positions) can be learned with reasonable success, these algorithms struggle to learn with multiple tasks at the same time, even with as few as ten distinct training tasks. Our analysis and open-source environments pave the way for future research in multi-task learning and meta-learning that can enable meaningful generalization, thereby unlocking the full potential of these methods.

Subfields: Artificial Intelligence, Cognitive Neuroscience, Reinforcement Learning, Evaluation & Benchmarks, 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

  1. Sergey Levine
  2. Chelsea Finn
  3. Yoshua Bengio
  4. Pieter Abbeel
  5. Mohit Bansal
  6. Ruslan Salakhutdinov
  7. Trevor Darrell
  8. Quoc V. Le

In Cognitive Neuroscience

  1. Sergey Levine
  2. Chelsea Finn
  3. Pieter Abbeel
  4. Yoshua Bengio
  5. Abhishek Gupta
  6. Shixiang Shane Gu
  7. Aviral Kumar
  8. Tianhe Yu

In Reinforcement Learning

  1. Sergey Levine
  2. Pieter Abbeel
  3. Chelsea Finn
  4. Ofir Nachum
  5. Shixiang Shane Gu
  6. Abhishek Gupta
  7. Abhinav Kumar Gupta
  8. Vikash Kumar

In Evaluation & Benchmarks

  1. Sergey Levine
  2. Jitendra Malik
  3. Abhinav Kumar Gupta
  4. Pieter Abbeel
  5. Silvio Savarese
  6. Trevor Darrell
  7. Chelsea Finn
  8. Dhruv Batra

In Meta & Continual Learning

  1. Chelsea Finn
  2. Eric Eaton
  3. Jorge Armando Mendez Mendez
  4. Sergey Levine
  5. Massimo Caccia
  6. Irina Rish
  7. Han-Jia Ye
  8. Razvan Pascanu

Citations in the corpus