Which one works faster? a) or b)

tensor = torch.randn((1, 2), requires_grad=True)

print(torch.tanh(tensor)) #a)
print(nn.Tanh()(tensor)) #b)

#nn.functional.tanh is deprecated
#print(torch.nn.functional.tanh(tensor))

They are actually the same, and so should be identical it terms of speed
Is you look at the source, nn.Tanh uses torch.tanh in its forward method.

I saw this just a moment ago. Thanks, I would say torch.tanh(tensor) is just a tiny bit faster. :wink:

BTW, do you have any idea what is @weak_script_method suppose to do?