I am fairly new to the Vision and Create ML framework. I downloaded the DetectingHumanActionsInALiveVideoFeed Project (https://developer.apple.com/documentation/createml/detecting_human_actions_in_a_live_video_feed#overview), and ran it on my iPhone 12 Pro Max and iPad Pro (fourth generation) with about 30 frames a second. When I tried running the app on an iPad (seventh generation) in an iPhone 7, I was only getting about five frames per second. Is there anyway to optimize that app to run on the A10 fusion chip? If it helps, all I need is for my app to read actions only. I don’t need any camera preview. Thanks in advance!