Comment by dirtbag__dad

This is also my experience, but when I saw how few tokens/second with a large model it gets locally I was bummed. I’m wondering if I’ll be obligated to upgrade in the near future