Train a balance bot with reinforcement learning

Shawn Hymel published a guide in Make: Volume 99 for training a two-wheeled balance bot with reinforcement learning using proximal policy optimization (PPO), the algorithm OpenAI published in 2017, rather than a classical control loop. PPO remains in use for robot control and for fine-tuning large language models.

Shawn Hymel wrote a guide in Make: Volume 99 on training a two-wheeled balance bot with reinforcement learning instead of a classical control loop. The agent is trained with proximal policy optimization, or PPO, the algorithm OpenAI published in 2017 that is still used for robot control and for fine-tuning large language models. Training was …