syaoran036 / pderl

Proximal Distilled Evolutionary Reinforcement Learning

Proximal Distilled Evolutionary Reinforcement Learning

This is modified version for PDERL algorithm.

original repo: https://github.com/crisbodnar/pderl

@inproceedings{bodnar2020proximal,
  title={Proximal distilled evolutionary reinforcement learning},
  author={Bodnar, Cristian and Day, Ben and Li{\'o}, Pietro},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={34},
  number={04},
  pages={3283--3290},
  year={2020}
}

UPDATE LOG

• update dependency gym to gymnasium

Run PDERL

install dependencies:

pip install -r requirements.txt

To run PDERL with proximal mutations and distillation-based crossover use:

python run_pderl.py -env=$ENV_NAME$ -distil -proximal_mut -mut_mag=$MUT_MAG$ -logdir=$LOG_DIR$

To evaluate and visualise a trained model in an environment use:

python play_pderl.py -env=$ENV_NAME$ -model_path=$MODEL_PATH$ -render 

ENVS TESTED

'Hopper-v4' \