Here is the original question by @luancordeiro:
Is there a built-in way to run cross-validation (e.g., K-fold or time-blocked) in EasyHybrid? If not, could you share a minimal Julia example showing how to set this up—rebuilding the model per fold, training on the train split, and evaluating on the holdout? Any quick tips on seeding/shuffling and batch norm during CV would also help.