/bmpo

Implementation of ICML2020 paper <Bidirectional Model-based Policy Optimization>

Primary LanguagePythonMIT LicenseMIT

Bidirectional Model-based Policy Optimization

This is the TensorFlow implementation for the paper Bidrectional Model-based Policy Optimization.

Requirements

pip install -r requirements.txt

Run

python main.py --config=config.hopperNT

To change hyper-parameters, please modify the corresponding config file in config/.

Acknowledgments

This code is mainly modified based on the mbpo codebase.