SQL Server 2016

By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.

Declare @YourTable Table ([AttendanceDate] date,[Col2] varchar(50),[Col3] varchar(50),[Col4] varchar(50),[Col5] varchar(50))  Insert Into @YourTable Values 
 ('2021-01-01','ABC','DEF',NULL,NULL)
,('2021-01-02','GH','ABC','DEF',NULL)
,('2021-01-03','GH','ABC','DEF',NULL)
,('2021-01-04','ABC','GH',NULL,NULL)
,('2021-01-05','GH',NULL,NULL,NULL)
,('2021-01-06',null,null,NULL,NULL)
,('2021-01-07','DEF',NULL,NULL,NULL)
,('2021-01-08','zz','DEF',NULL,NULL)
,('2021-01-08',null, null,NULL,NULL)
,('2021-01-08','xx', null,NULL,NULL)
,('2021-01-08','xx', null,NULL,NULL)
,('2021-01-09','xx', null,NULL,NULL)
,('2021-01-10','xx', null,NULL,NULL)

;with cte as (
Select AttendanceDate
      ,B.*
      ,Grp = datediff(day,0,AttendanceDate) - dense_rank() over (partition by value order by AttendanceDate)
 From  @YourTable A
 Cross Apply (
                Select [Key]
                      ,Value
                 From OpenJson((Select A.* For JSON Path,Without_Array_Wrapper )) 
                 Where [Key] not in ('AttendanceDate')
             ) B
)
Select [Value]
      ,[Start]  = min(AttendanceDate)
      ,[End]    = max(AttendanceDate)
 From  cte
 Group By Value,Grp

Value	Start	End
ABC	2021-01-01	2021-01-04
DEF	2021-01-01	2021-01-03
GH	2021-01-02	2021-01-05
DEF	2021-01-07	2021-01-08
xx	2021-01-08	2021-01-10
zz	2021-01-08	2021-01-08