If you use this code, please cite the following paper
Anirudh Vemula, Wen Sun, J. Andrew Bagnell. Contrasting Exploration in Parameter and Action Space: A Zeroth Order Optimization Perspective. In the proceedings of Artificial Intelligence and Statistics (AISTATS). 2019
Run
pip install -r requirements.txt
./scripts/setup.sh
To reproduce the results of the Mujoco Experiments (both Swimmer and HalfCheetah) with hyperparameter tuning, run
./scripts/run_mujoco_experiments.sh
To reproduce the results of the LQR experiments with hyperparameter tuning, run
./scripts/run_lqr_experiments.sh
To generate plots used in the paper, run the following commands after running the experiments
python scripts/plot_zooomed_results.py
python scripts/plot_lqr_results.py
- Python 3.6.3
- Ray 0.5.2
- Numpy 1.15.1
- Matplotlib 2.2.3
- Mujoco-py 1.50.1.56
- Gym 0.10.5
- Scipy 1.1.0
- Jupyter 1.0.0
- Autograd