I'm not sure what you are asking. Sure you write the magnitude (in Newtons or whatever) if you know it. Otherwise, write it as a variable and try to solve for it later.
Magnitude as in the "length" or norm of the vector i.e. 5N, or 4m/s.
So, there's no convention that forces you to write the magnitude if you know it at the beginning of a problem?
One is usually better off using symbols for known and unknown quantities alike, and using algebra to express unknowns via knowns. The numbers can be plugged in at the very end.