unable to use inverse_transforms in infers

Can't add any code to inverse_transforms gives error

class MySegmentation(BasicInferTask): 
....    
def inverse_transforms(self, data=None) -> Union[None, Sequence[Callable]]:
return [
RestoreOrientationd(keys="pred", ref_image="image")
]

gives error

 File "/MONAILabel/monailabel/tasks/infer/basic_infer.py", line 306, in __call__
    data = self.run_invert_transforms(data, pre_transforms, self.inverse_transforms(data))
  File "/MONAILabel/monailabel/tasks/infer/basic_infer.py", line 397, in run_invert_transforms
    transforms = [pre_transforms[n if isinstance(n, str) else n.__name__] for n in names]
  File "/MONAILabel/monailabel/tasks/infer/basic_infer.py", line 397, in <listcomp>
    transforms = [pre_transforms[n if isinstance(n, str) else n.__name__] for n in names]
AttributeError: 'Orientationd' object has no attribute '__name__'

the only way I can get it to work is to create a fake class and add attribute name as

    def inverse_transforms(self, data=None) -> Union[None, Sequence[Callable]]:
        class CodeHack:
            pass
        codehack = CodeHack()  # Create an empty employee record
        codehack.__name__ = 7

        return [codehack]

this would invert the transform number 7 without any errors.

Start in monailabel/tasks/infer/basic_infer.py at BasicInferTask.call and run_invert_transforms, around lines 306 and 397. Trace how inverse_transforms handles RestoreOrientationd and how transform names are resolved. Done means the reported AttributeError no longer occurs when using RestoreOrientationd in inverse_transforms and the transform is inverted as intended.