MT-Opt: Continuous Multi-Task Robotic Reinforcement Learning at Scale

Builds on top of QT-Opt.

MT-opt = multi-task optimization?

Project page:

• https://karolhausman.github.io/mt-opt/

Citation

(Kalashnikov et al., 2021)

Kalashnikov, D., Varley, J., Chebotar, Y., Swanson, B., Jonschkowski, R., Finn, C., Levine, S., & Hausman, K. (2021). MT-Opt: Continuous Multi-Task Robotic Reinforcement Learning at Scale. https://arxiv.org/pdf/2104.08212