0

i build a classifying model about phishing site.

1. first, about my data

num_dataset: i have about 16000 dataset

num_feature: my dataset has 12 characterisitics.

label: if it's a phishing site, i set it's label -1. else 1

batch_size: set batch_size 128, for cnn model

kernel_size: 3

2. I try

from torch.utils.data import DataLoader, TensorDataset

train_dataset = TensorDataset(x_train, y_train)
train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True)

#-------------------------------------------------------#
class CNN(nn.Module):
  def __init__(self, kernel_size):
    super().__init__()  
    self.layer1 = nn.Sequential(
        nn.Conv1d(12, 32, 3,
                  stride=1, padding=1)
...

  def forward(self, x):
    out = self.layer1(x)
...
#-------------------------------------------------------#
model = CNN()

for epoch in range(epochs):
  avg_cost = 0

  for x_train, y_train in train_loader:

    Hypothesis = model(x_train)

x_train.shape

torch.Size([16072, 12])

y_train.shape

torch.Size([16072, 1])

3. Error

at Hypothesis = model(x_train)

RuntimeError: Expected 3-dimensional input for 3-dimensional weight [32, 12, 3], but got 2-dimensional input of size [128, 12] instead

4. Finnally

i think it's because I'm confused between conv1d and conv2d but i can't figure it out...

plz, i want to know the cause of this problem

UiJin
  • 175
  • 1
  • 1
  • 6

1 Answers1

1

You are using a nn.Conv1d which should receive a 3-dimensional input shaped (batch_size, n_channels, sequence_length). This being said your input has n_channels=12 (since you've initialized your 1d conv with 12 input channels) and a sequence_length=1. To match the requirements, you need to have an additional dimension on your input. Passing something like x_train.unsqueeze(-1) to your network should work.

See the following example:

>>> x = torch.rand(100, 12, 1)
>>> L = nn.Conv1d(12, 32, 3, stride=1, padding=1)
>>> L(x).shape
torch.Size([100, 32, 1])
Ivan
  • 34,531
  • 8
  • 55
  • 100