Dynamically mocking the return value of a database call in Django test

I am trying to test a function and dynamically mock out the database call to the RecordType table, as the test database does not contains all the instances that I need for the function to run. This is the part of the function in question:

    permanent = []
    temp = []
    for category in enums.Category:
        category_name = str(category.name)
        if category_name in labels:
            if not category.is_temporary_code:
                permanent.append(category)
            else:
                temp.append(category)

    for category in permanent:
        category_name = str(category.name)
        internal_code = models.RecordType.objects.get(
            industry_code=category.value
        ).internal_code
        field_label = labels[category_name]
        self.fields[internal_code] = forms.BooleanField(label=field_label, required=False)

I have a dict of values (record_type_dict) that I would like to be returned in place of calling the mocked function, depending on the passed in 'category.value'. In my test file I have this code:

@mock.patch.object(forms, 'models')
def test_records(mock_models, factory, support_client):

....

mock_models.RecordType.objects.get().internal_code.side_effect = side_effect

...


def side_effect(*args, **kwargs):
    if args[0] in record_type_dict:
        return record_type_dict[args[0]]
    else:
        return ""

However this does not work as the function is being called with 'industry_code=category.value'.

Is there a way of dynamically mocking this call depending only on the value of category.value?