How many test samples are included in each train-test split when using a 20% test size?

How many test samples are included in each train-test split when using a 20% test size?: a task in SmolDataEnvs-harbor-train-sandbox: 📊 SmolDataEnvs: Harbor (train) (Harbor dataset). Files (in /home/user/input, no subfolders): - Iris.csv