Puppo the Corgi is a Unity ML-Agents Toolkit machine learning mobile game project where you play fetch with a cute little corgi. Throw a stick to Puppo by swiping on the screen and Puppo brings it back. I used the ML-Agents Toolkit to train my own version of Puppo, and extended the original demo to add in a few gameplay features, setup a full gameplay loop, and get a working Android build that is available for testing below. This is a work in progress and will continue to be extended with further features and levels:
- AI – Used ML-Agents Toolkit to perform reinforcement machine learning to train my own version of Puppo.
- Core – Programmed game manager and loading screen systems. Setup gameplay loop and created Android demo build.
- Engine – Reorganize and maintain project hierarchy, folder structure, game settings, and engine settings to keep all parts of project modular. Performance and optimization improvements.
- Gameplay – R Setup animal color picker (work in progress). Change Puppo’s model to a llama model for Lexi the Llama version of game.
- QA and Systems – Responsible for testing, debugging, tuning, balancing, and git source control.
- Android Demo Available – Demo available now only on Android: Download Demo