flow-to-better

Official code for FTB: Flow to Better: Offline Preference-based Reinforcement Learning via Preferred Trajectory Generation.

Usage

First, you need to install Python packages listed in requirements.txt using pip install -r requirements.txt.

Then, just run the scripts in the run_scripts dir. You can try as follows:

For mujoco tasks:

sh run_scripts/run_mujoco.sh

For Metaworld tasks:

sh run_scripts/run_metaworld.sh

For NeoRL tasks:

sh run_scripts/run_neorl.sh

Citation

If you find this repository useful for your research, please cite:

@inproceedings{
	zhang2024flow,
	title={Flow to Better: Offline Preference-based Reinforcement Learning via Preferred Trajectory Generation},
	author={Zhilong Zhang and Yihao Sun and Junyin Ye and Tian-Shuo Liu and Jiaji Zhang and Yang Yu},
	booktitle={The Twelfth International Conference on Learning Representations},
	year={2024},
	url={https://openreview.net/forum?id=EG68RSznLT}
}