Q-Learning - Frozen Lake 8x8 Enhanced
Welcome to the FrozenLake8x8-DQL-Enhanced repository! This project is designed to provide an enhanced learning experience for understanding Q-Learning in the FrozenLake-v1 environment. The enhancements include visualizing Q values in real-time, enlarging the map for better readability, and incorporating shortcut keys for animation control.
1. Real-time Q Value Visualization: The Q values are dynamically overlaid on each cell of the map. This allows you to witness the Q values update in real-time as the agent undergoes training, providing valuable insights into the learning process.
2. Enlarged Map Display: The map is enlarged to fill the entire screen, ensuring that the overlaid Q values are easier to read. This enhancement improves the overall visibility and clarity of the learning environment.
3. Shortcut Keys for Animation Control: To facilitate a smoother learning experience, shortcut keys have been implemented to control the animation speed. You can speed up or slow down the training animation based on your preference.
1. frozen_lake_qe.py: This file is nearly identical to the original frozen_lake_q.py, with the key difference being the utilization of the frozen_lake_enhanced.py environment. Use this script to execute Q-Learning in the enhanced FrozenLake-v1 environment.
2. frozen_lake_enhanced.py: This module provides the FrozenLake-v1 environment overlaid with Q values. You don't need an in-depth understanding of this code; it serves as the foundation for visual enhancements.
Follow these steps to get started with the FrozenLake8x8-DQL-Enhanced project:
1. Clone the repository to your local machine.
git clone https://github.com/your-username/FrozenLake8x8-DQL-Enhanced.git
cd FrozenLake8x8-DQL-Enhanced
2. Run the Q-Learning algorithm using the enhanced environment.
python frozen_lake_qe.py
or Just press F5 to open debug Terminal
3. Explore and modify the code as needed, and enjoy the visualized Q-Learning experience!
This project builds upon the work from JohnnyCode8's Gym Solutions. We would like to express our gratitude for their contributions and inspiration to enhance the Frozen Lake 8x8 environment for Q-Learning.
This project is licensed under the MIT License - see the LICENSE.md file for details.
Feel free to contribute, provide feedback, or report issues. Happy learning with Q-Learning in the enhanced Frozen Lake 8x8 environment!