The promise of Evolutionary Robotics (ER) to completely automatize the design of robot controllers and/or morphologies is an idea with great appeal not only to researchers, but also to students. However, when attempting to grab and hold student interest, the large requirements of time and computational resources required to achieve good results in ER systems may be discouraging. In fact, after two years of using our RoboGen evolutionary robotics system for class projects, the biggest student complaints all concerned the slow speed of evolutionary progress. In order to overcome these limitations, we investigate a simple and effective technique for rapidly evolving robot gaits in a manner of seconds or minutes rather than hours or days. We rely on two basic techniques to speed up evolution: Compositional Pattern Producing Network (CPPN) encodings and simple parameterized oscillator neurons. When combined with a previously executed iterative tuning procedure, many of these evolved gaits can be transferred to real robots with reasonable fidelity.