I would like to know how experienced programmers and software engineers balance usage of Django's fat models approach against DRF views. In practice, my experience revolves around configuring DRF serializers and views more than creating and overriding helper functions inside model classes. It may help to mention that I prefer to keep models loose for easier modification (e.g. creation logic at the API level vs at the model level).
As you can imagine, my serializers and API views contain the custom logic and I don't know if this is expected since these DRF components are at the frontline and not the model methods that are primarily suited, for example, for core Django apps/components.
Disclaimer: I am no expert in Django and DRF.