This is the official PyTorch implementation of the paper "Boosting Continuous Control with Consistency Policy". For those interested in delving deeper into our research, you can find detailed versions of our paper:

For an extended read, including the appendix, check out the Arxiv Version. For the conference-specific details as presented at AAMAS 2024, access the AAMAS 2024 Version.

git clone https://github.com/cccedric/cpql.git
cd cpql

conda env create -f cpql_env.yaml

• Install mujoco210 and mujoco-py following instructions here.
• Install D4RL following instructions here.

python main.py --rl_type offline --env_name hopper-medium-expert-v2

python main.py --rl_type online --env_name Hopper-v3

For any questions, please feel free to email [email protected].

Our code is built upon consistency models, Diffusion-QL. We thank all these authors for their nicely open sourced code and their great contributions to the community.

This repository is released under the GNU license. See LICENSE for additional details.

If you find our research helpful and would like to reference it in your work, please consider using one of the following citations, depending on the format that best suits your needs:

For the Arxiv version:

@article{chen2023boosting,
  title={Boosting Continuous Control with Consistency Policy},
  author={Chen, Yuhui and Li, Haoran and Zhao, Dongbin},
  journal={arXiv preprint arXiv:2310.06343},
  year={2023}
}

Or, for citing our work presented at the conference of AAMAS 2024:

@inproceedings{chen2023boosting,
  author={Yuhui Chen and Haoran Li and Dongbin Zhao},
  title={Boosting Continuous Control with Consistency Policy},
  booktitle={Proceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems, {AAMAS} 2024, Auckland, New Zealand, May 6-10, 2024},
  pages={335--344},
  publisher={ACM},
  doi={10.5555/3635637.3662882},
}