DMRG CASPT2 Problem with active space

Hello Everyone,
I have a problem with placing the appropriate orbitals in the active space for the compound with transition metal. The idea is to place the deeply lying d and pi orbitals in the active space. Could someone explain it to me technically? Or maybe the only option is to use the keyword WhichActiveSpace_in? If so, how to use it correctly? Maybe can these orbital be set manually in input?
Thank you in advance for your answer and I am asking for your understanding (I am just starting my adventure with Psi4 program package)
Yours Sincerely,

Apologies for the slow response. DMRG is a add-on and not directly related to Psi4. You may get some answers here or you can try on the DMRG website. Pinging @francesco