1MV Cash Application Machine Learning Model - How to get 5,000-10,000 cleared payments in one Model

Dear Sir oregon Madam,

The pursuing question has arisen erstwhile utilizing instrumentality learning to post-process the slope statement.

This documentation https://me.sap.com/notes/0003001671 is pointing retired that 5,000 -10,000 cleared payments indispensable beryllium disposable successful bid to bid the ML exemplary sufficiently.

To bash this, we transportation retired the pursuing steps:

1. ML_CASH_APP_JOB_STATUS_DP and acceptable presumption to “In Progress” (P_INPROG)
2. ML_CASH_APP_DATA_POST to extract grooming data
3. ML_CASH_APP_JOB_STATUS_DP and acceptable presumption to “Complete” (P_COMPL) -> This volition trigger grooming successful cloud

The Model Manager for SAP Cash Application tin present beryllium recovered successful the BTP Cockpit:

https://help.sap.com/docs/SAP_CASH_APPLICATION/9668f05979b644cd960f0d83ae4ed0aa/73fd86f7258b46bf84ef...

When steps 1 to 3 are executed, a caller inactive exemplary is created. With respect to the processed items, the statistic amusement that not each cleared payments from the ERP are processed cumulatively. We tin admit this by the fig of items erstwhile we tally this grooming occupation each month. The fig does not accumulate. In addition, each grooming tally creates a caller exemplary that indispensable past beryllium activated.

How are we expected to get an ML exemplary that has generated 5,000-10,000 cleared payments if each grooming tally lone processes a portion oregon caller cleared payments and past generates a caller exemplary connected this basis?

Who tin explicate this successful detail?

Thank you precise much
Sebastian