Delete duplicate rows (don't delete all duplicate)

I am using postgres. I want to delete Duplicate rows. The condition is that , 1 copy from the set of duplicate rows would not be deleted.

i.e : if there are 5 duplicate records then 4 of them will be deleted.