Reinforcement Learning Project with Stable-Baseline3 and Gym

Overview

This project demonstrates the implementation of reinforcement learning algorithms using Stable-Baseline3 library in Python. The environment for training and testing the RL agents is provided by OpenAI Gym.

Dependencies

Ensure you have Python 3 installed on your system. You can install the required dependencies using pip and the provided requirements.txt file:

pip install -r venv/requirements.txt

This will install all the necessary packages including Stable-Baseline3, Gym, and other dependencies.

Running the Code

All the code is contained within the main.ipynb Jupyter notebook. You can open and run this notebook using Jupyter Notebook or JupyterLab:

jupyter notebook main.ipynb

This notebook contains code for training RL agents, testing them, and evaluating their performance.

TensorBoard Integration

To visualize the training progress and monitor various metrics, you can use TensorBoard. We've provided a shell script tensorboard.sh for convenience. Simply run:

sh tensorboard.sh

This will start TensorBoard server and you can access it through your browser at http://localhost:6006.

Additional Notes

• Ensure you have a compatible CUDA version installed if you're planning to utilize GPU acceleration.
• Feel free to explore different RL algorithms provided by Stable-Baseline3 and experiment with various environments in Gym.
• Adjust hyperparameters and tweak the code to suit your specific use case and environment.

Credits

This project utilizes the following libraries and resources:

• Stable-Baseline3: GitHub Repository
• OpenAI Gym: GitHub Repository

License

This project is licensed under the MIT License.