No, you can run “cog predict” and test on your cloud GPU before pushing.

→ View original post on X — @fofrai