Renhao Wang, Kevin Frans, Pieter Abbeel, Sergey Levine, Alexei A. Efros
This repository is based heavily off the release code of SynthER. The instructions for setting up their environment are reproduced below:
To install, clone the repository and run the following:
git submodule update --init --recursive
pip install -r requirements.txtOur code is tested on Python 3.8. If you don't have MuJoCo installed, follow the instructions here: https://github.com/openai/mujoco-py#install-mujoco.
python synther/online/online_cond.py --env quadruped-walk-v0 --gin_config_files config/online/sac_cond_synther_dmc.gin --gin_params 'redq_sac.cond_top_frac = 0.1'python synther/online/online_cond.py --env cheetah-run-v0 --gin_config_files config/online/sac_cond_synther_dmc.gin --gin_params 'redq_sac.cond_top_frac = 0.25'python synther/online/online_cond.py --env reacher-hard-v0 --gin_config_files config/online/sac_cond_synther_dmc.gin --gin_params 'redq_sac.cond_top_frac = 0.25'@inproceedings{wang2025prioritized,
title={Prioritized Generative Replay},
author={Renhao Wang and Kevin Frans and Pieter Abbeel and Sergey Levine and Alexei A Efros},
booktitle={The Thirteenth International Conference on Learning Representations},
year={2025},
url={https://openreview.net/forum?id=5IkDAfabuo}
}This codebase inherits all licenses from the public release of SynthER.