How to implement the GAT model to a regression problem? **Particularly the design of labels**

[SuperBruceJia]

Dear,

How can I implement the GAT model to a regression problem?

As the labels of regression task are between an interval, e.g., 0 - 100.
Can I simply connect the output of Multi-head attention with a few FC layers, and finally a single neuron, i.e., FC layer with one neuron to predict and calculate the loss between the output and label?

I'm looking forward to your early reply.

Best,

Shuyue