There are many ways to delete the duplicates in a table but sometimes the group by doesnt seem to work so you can try out the new feature of Microsoft SQL Server "Common Table Expression"
With Dups as
SELECT row_number() over (
partition by ColumnName
order by ColumnName ) as RowNum
WHERE rownum > 1