tensorflow - multi output multi label - same loss function?

Question

ok so i have a model that outputs 2 predictions e.g height of a person and some properties of the population so i have:

OUTPUT:

(Batch,1) (one prediction for