Hi, I have a question about training accuracy vs F1-score in FOMO. I’ve done an object recognition project, and as I can see, when I do some versioning of my project, it seems that the training accuracy on the version page is a rounded value of the F1-score. Is that correct?