$[$ From in$-$class discussion: weeks $11$ & $12]$ Which statement is FALSE regarding Federated Machine Learning $($FML$)$:

In FML$,$ each node trains a model locally and sends only the updated model parameters to a global model hosted in the cloud.

FML enables multiple devices$/$servers $($nodes$)$ to learn a shared model collaboratively without divulging their raw data.

In the FML approach, an aggregator $($aka: server or coordinator$)$ facilitates the multi$-$party collaboration.

Once a global model is trained by FML clients, the participating clients all share the global model via the cloud.