Cooperative training methods for distributed machine learning typically\nassume noiseless and ideal communication channels. This work studies some of\nthe opportunities and challenges arising from the presence of wireless\ncommunication links. We specifically consider wireless implementations of\nFederated Learning (FL) and Federated Distillation (FD), as well as of a novel\nHybrid Federated Distillation (HFD) scheme. Both digital implementations based\non separate source-channel coding and over-the-air computing implementations\nbased on joint source-channel coding are proposed and evaluated over Gaussian\nmultiple-access channels.\n