Initialize weights to CNN

[Flock1 (Flock Anizak)]

Hello,

I have some initial weights that I want to initialize a CNN layer with. The weights are of the shape (3,8,12). I defined a CNN layer as

conv1 = nn.Conv2d(3,8,kernel_size=(3,3), stride = 2)

I just guessed this initialization and that’s why I am getting the following error:

RuntimeError: Given weight of size [1, 3, 8, 12], expected bias to be 1-dimensional with 1 elements, but got bias of size [8] instead

My input is of size (3,244,244). Kindly let me know how can I initialize the CNN layer so that it accepts the weights.

[Programmer-RD-AI]

Hi,
Can you check

RuntimeError: Given weight of size 3 1 5 5, expected bias to be 1-dimensional with 3 elements, but got bias of size [6] instead

Your model seems to work fine with bias=False and bias=True: class LeNet(nn.Module): def __init__(self): super(LeNet, self).__init__() self.conv1 = nn.Conv2d(1, 6, 5,bias=False) self.conv2 = nn.Conv2d(6, 16, 5) self.fc1 = nn.Linear(16 * 5 * 5, 120) self.fc2 = nn.Linear(120, 84) self.fc3 = nn.Linear(84, 10) def forward(self, x): out = self.conv1(x) out = F.relu(out) out = F.max_pool2d(out, 2) out = F.rel…

RuntimeError: Given groups=1, weight[64, 3, 3, 3], so expected input[16, 64, 256, 256] to have 3 channels, but got 64 channels instead

Your input contains 4 channels, while the first conv layer expects an input with 3 channels. If you are dealing with RGB images, the 4th channel might be the alpha channel, which could just be removed. If you are using a custom DataLoader, you could probably just use: def __getitem__(self, index): img = Image.open(self.paths[index]).convert('RGB') ... # Alternatively remove the alpha channel from the tensor img = Image.open(self.paths[index]) x = TF.to_tensor(img) x = …

if the above doesn’t work
can you check the links below

https://www.google.com/search?q=RuntimeError%3A+Given+weight+of+size+[1%2C+3%2C+8%2C+12]%2C+expected+bias+to+be+1-dimensional+with+1+elements%2C+but+got+bias+of+size+[8]+instead&oq=RuntimeError%3A+Given+weight+of+size+[1%2C+3%2C+8%2C+12]%2C+expected+bias+to+be+1-dimensional+with+1+elements%2C+but+got+bias+of+size+[8]+instead&aqs=chrome..69i57j69i58&sourceid=chrome&ie=UTF-8

With best regards,
Programmer-RD-AI

[my3bikaht (Sergey)]

@Flock1

conv1 = nn.Conv2d(3,8,kernel_size=(3,3), stride = 2)

Not sure where 12 came from. Your conv1 layer has 3 input channels, 8 output channels, kernel size of 3x3 and stride 2. Thus, you have 8 kernels, each has 3x(3x3) weights and 1 bias

Use conv1.bias.shape and conv1.weight.shape after initialization to get correct dimensions.