from torchinfo import summary
h, w = 128, 128
print(f"Expected total multiplies: {m * dk * dk * h * w * n}")
summary(conv1, input_size=(1, m, h, w))