Got 32GB so was hoping I could use ollm to offload it to my SSD. Slower but making it possible to run bigger models (in emergencies)