CoTASP

Code for "Continual Task Allocation in Meta-Policy Network via Sparse Prompting", presented in ICML 2023.

Key Dependencies

The some of dependencies are outdated, e.g., the ones listed below. You might need to install their latest version to run this project.

python==3.7.13
- jax==0.3.17
- jaxlib==0.3.15+cuda11.cudnn82
- flax==0.6.4
- optax==0.1.4
- scikit-learn==1.0.2
- tensorflow-probability==0.18.0
- sentence-transformers==2.2.2

Refer to this repo for the installation of Continual World.

Quick Start

python train_cotasp.py

Reproducibility

Tracked experiments on CW20 via Weights & Biases.

Citing CoTASP

If you use the code in CoTASP, please kindly cite our paper using following BibTeX entry.

@InProceedings{pmlr-v202-yang23t,
  title = 	 {Continual Task Allocation in Meta-Policy Network via Sparse Prompting},
  author =       {Yang, Yijun and Zhou, Tianyi and Jiang, Jing and Long, Guodong and Shi, Yuhui},
  booktitle = 	 {Proceedings of the 40th International Conference on Machine Learning},
  pages = 	 {39623--39638},
  year = 	 {2023},
  volume = 	 {202},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {23--29 Jul},
  publisher =    {PMLR},
  pdf = 	 {https://proceedings.mlr.press/v202/yang23t/yang23t.pdf},
  url = 	 {https://proceedings.mlr.press/v202/yang23t.html},
}

Acknowledgement

We appreciate the open source of the following projects:

Continual World, Meta World, and JaxRL