Hi! I have been working on an object detector using the Create ML app and I was confused about the metrics it gave me, as well. I/U means "Intersection over Union". It is a very simple metric which tells you how good the model is at predicting a bounding box of an object. You can read more about I/U in the "Intersection over Union (IoU) for object detection" article at PyImageSearch. So: I/U 50% means how many observations have their I/U score over 50% Varied I/U is an average I/U score for a data set Note: it is my understanding of Apple's intention of giving those metrics. What I mean is I am not exactly sure if they meant > or >= 50% in the first case and if it really is an average in the second case, but I am pretty sure it is just the way it works. Take care!

@mellis94, the crash happens only in the Xcode preview. When running in the simulator or on a physical device, everything is fine.