If I get a low evaluation score for a ML Package, and I want to start completely over with a cleaned up training dataset – is there an easy way to “reset” everything? Or must I delete the package and recreate it completely? And delete all the pipelines one by one? Or is there a way to just go back the beginning version and delete all the minor versions after the first one?
If you mean you want to use a new dataset, you can just create a new dataset and train the model with the base minor version (eg: x.x.0). It will always train to become a new model.
Regards,
Azeem