Is there any difference between these 2 usage which is shown in below. I mean in the way of computational graph or any difference which is realted to backward?
def __init__(self, input_dim, hidden_dim, batch_size)
super(model, self).__init__()
self.softmax = F.softmax
def forward():
...
out = self.softmax(out) vs out = F.softmax(out)