Pytorch tutorial max pooling reuse

vision
1

I transferred from TensorFlow recently. In this tutorial, there is a snippet:

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(3, 6, 5)
        self.pool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(6, 16, 5)

    def forward(self, x):
        x = self.pool(F.relu(self.conv1(x)))
        x = self.pool(F.relu(self.conv2(x)))
        return x

I have two questions:

  1. Will the max pooling in two different layers share the same weights?
  2. If not, how would PyTorch know they are different?

Thank you in advance.

2

That is because MaxPool2d has no trainable parameters.
image
see, output nothing.

2 Likes
3

How did I forget that!