Model version control for native ML Learners in Pipeline Pilot

Hello! I am after some guidance best practices to manage version control on the learners. I have a protocol that is run using the scheduler (every two weeks linked to a evolving database). Every new version is always overwriten as version #1 when I would like to have all the versions ( 1,2, ... ) and the corresponding time. I haven't found a way how to enable the version 1,2,3, ... in the trained model component. Is it expected that we need to save the new model with for example the date of training as a way to keep track of versions. Any guidance on best pratice with ML Learners is welcome. Thanks.