M L Agents Pyramids

Introduction

The MLAgents-Pyramids project involves a Proximal Policy Optimization (PPO) agent trained using the Unity ML-Agents library. This model is designed to play the Pyramids game, showcasing the capabilities of reinforcement learning.

Architecture

The model utilizes the Unity ML-Agents framework, which supports the development and training of reinforcement learning models. The PPO algorithm is employed for training the agent, providing an effective policy gradient approach for continuous control tasks.

Training

To train the PPO agent using the ML-Agents library, users must configure their environment and training parameters. Training can be resumed using the mlagents-learn command with the appropriate configuration file and run ID. This flexibility allows users to continue improving the agent's performance over multiple training sessions.

Guide: Running Locally

1. Set Up Environment:

   • Install the Unity ML-Agents toolkit from the GitHub repository.
   • Ensure you have the necessary dependencies installed, such as Python and the TensorFlow library.

2. Resume Training:

   • Use the command mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume to resume training from a saved checkpoint.

3. Watch the Agent Play:

   • Visit the Hugging Face Space.
   • Enter the model ID: ThomasSimonini/MLAgents-Pyramids.
   • Select the appropriate .nn or .onnx file and click on "Watch the agent play."

4. Cloud GPU Suggestion:

   • For enhanced performance, consider using cloud GPU services like AWS, Google Cloud, or Azure to handle computationally intensive tasks.

License

The project is shared under the terms specified by the original repository and any additional licenses contained within related libraries or frameworks. Users should review these licenses to ensure compliance.