The project is designed to simulate hierarchical reinforcement learning algorithms. There is two various environments: grid_maze_env README and arm_env README. One should check the environments' READMEs for more information.
You can run experiments with handcrafted machines hierarchies in module ham_experiments. And also examine HAM's readme file.
Things you need to install the software:
sudo apt-get install python3-tk
sudo apt-get install python3-dev
For drawing graphs with pygraphviz one should install:
sudo apt-get install graphviz
sudo apt-get install graphviz-dev
sudo apt-get install python3.5-dev
sudo pip3.5 install pygraphviz --install-option="--include-path=/usr/include/graphviz" --install-option="--library-path=/usr/lib/graphviz/"
To run random_policy.py and test the environment, you must install the following libraries:
gym
scipy
pandas
matplotlib
numpy
pygraphviz
Run the file q-policy.py, which will show an example of interaction on both environments with q-learning and random policy.
- Alexander Panov - Project management - grafft
- Alexey Skrynnik - Environments. Hierarchical RL on HAMs - Tviskaron
- Vadim Kuzmin - Hierarchical RL on Options - vtkuzmin
This project is licensed under the Apache License 2.0 - see the LICENSE file for details