Let\'s say I have 8 cores, and I want to run 2 functions that each use 4 cores (ML models for example), what is the best way to set this up. I cannot find examples such as this