Mean() vers min()

Hi,

I have a basic question from chapter 4, calculating the tensor mean along dimension 0.
The code from the notebook is:
mean3 = stacked_threes.mean(0)
show_image(mean3)

Playing around with the code I changed it to:
min3 = stacked_threes.min(0)
show_image(min3)

However this results in an error:
ValueError: only one element tensors can be converted to Python scalars
Why is this??

The difference seams to be the object types returned from mean() and min(). They are <class ‘torch.Tensor’> and <class ‘torch.return_types.min’>. But why is this?
I was expecting it to take the min along dimension 0 of the stacked tensor of 3s.

thanks
Norbert