[taekop]

group_vfold_cv() hardcoded repeats = 1 in the attributes of its result, so attr(x, "repeats") was wrong when repeats > 1. It now records the supplied value, as vfold_cv() does.

library(rsample)
attr(group_vfold_cv(mtcars, cyl, v = 2, repeats = 2), "repeats")
#> [1] 1   # before
#> [1] 2   # after

Closes #607