MicroRacer: a didactic environment for Deep Reinforcement Learning
This provides a didactic tool for educators and learners in Deep Reinforcement Learning, but it is incremental as it builds on existing environments with a focus on simplicity and accessibility.
The authors introduced MicroRacer, a simple, open-source car racing environment designed for teaching Deep Reinforcement Learning, with calibrated complexity to enable quick experimentation with various methods and hyperparameters, and provided baseline agents for algorithms like DDPG and PPO with preliminary comparisons of training time and performance.
MicroRacer is a simple, open source environment inspired by car racing especially meant for the didactics of Deep Reinforcement Learning. The complexity of the environment has been explicitly calibrated to allow users to experiment with many different methods, networks and hyperparameters settings without requiring sophisticated software or the need of exceedingly long training times. Baseline agents for major learning algorithms such as DDPG, PPO, SAC, TD2 and DSAC are provided too, along with a preliminary comparison in terms of training time and performance.