There's some serious misunderstanding here. You do not have to disambiguate the function call, only the construction of the object. You would write m1 = Tens

There's some serious misunderstanding here. You do not have to disambiguate the function call, only the construction of the object. You would write m1 = TensorFlow.model() fit(m1, data) m2 = Pytorch.model() fit(m2, data) Julia knows which version of model you are using. YensorFlow.fit and Pytorch.fit are just different methods of the same function. You've formed some strong opinions based on a pretty big misunderstanding.