[DNF2]

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.