In PyTorch, how to make __call__() of nn.Module automatically copy the type hints and docstring of forward()?
It is common to implement the forward steps in forward() and call it through __call__, like this example
In PyTorch, how to make __call__() of nn.Module automatically copy the type hints and docstring of forward()?
It is common to implement the forward steps in forward() and call it through __call__, like this example