Sorting out duplicates after join

[Rmiddend]

Hello,

I am building out a system that takes two data streams and joins them together if identical to each other. The issue is that sometimes they are identical to each other but with differing reference numbers. This results in the output doubling, tripling, or worse. I have attached a flow with an example of the date with a start point and an ideal end point. I basically need to sort these in a way where only one record exists for a unique cell in field 2 and a unique cell in field 3.

Thanks for any help, im very stumped on this.

[Rmiddend]

Further on this, these numbers will change with each run so the solution needs to be able to do this with different numbers each time.

[Qiu]

@Rmiddend 
We better to check the Join tool rather than after the Join too, otherwise we have to decide the selection criteria for Field3.
If that does not matter, an unique tool shall do it.