Only CPU machine implementation

[Priyankamun]

I am trying to run the code on a cpu only machine. Please can anyone tell me the changes i need to make in all the codes. Are there any ways to execute on a NO GPU machine?

I have explored a few options but stuck right now at
runfile('C:/Users/Admin/.spyder-py3/Capsule-Forensics-v2-master/train_binary_ffpp.py', wdir='C:/Users/Admin/.spyder-py3/Capsule-Forensics-v2-master')
Namespace(batchSize=32, beta1=0.9, cpu_id=0, dataset='databases/faceforensicspp', disable_random=False, dropout=0.05, imageSize=300, lr=0.0005, manualSeed=None, niter=25, outf='checkpoints/binary_faceforensicspp', resume=0, train_set='train', val_set='validation', workers=0)
Random Seed: 3435
Traceback (most recent call last):

File "C:\Users\Admin.spyder-py3\Capsule-Forensics-v2-master\train_binary_ffpp.py", line 76, in
capsule_loss = model_big.CapsuleLoss(opt.cpu_id)

File "C:\Users\Admin.spyder-py3\Capsule-Forensics-v2-master\model_big.py", line 200, in init
self.cross_entropy_loss.to('cpu')(cpu_id)

File "C:\Users\Admin\Miniconda3\lib\site-packages\torch\nn\modules\module.py", line 493, in call
result = self.forward(*input, **kwargs)

TypeError: forward() missing 1 required positional argument: 'target'

[honghuy127]

Yes, just simply delete .cuda(...) in the scripts.

To load saved GPU models to CPU-only machine, just modify the load part as follow:
torch.load('my_file.pt', map_location=lambda storage, loc: storage)

[Priyankamun]

Thank you, I made the changes you mentioned. That got me here:

def forward(self, input, output_size=None):
# type: (Tensor, Optional[List[int]]) -> Tensor
if self.padding_mode != 'zeros':
raise ValueError('Only zeros padding mode is supported for ConvTranspose1d')

    output_padding = self._output_padding(input, output_size, self.stride, self.padding, self.kernel_size)
    return F.conv_transpose1d(
        input, self.weight, self.bias, self.stride, self.padding,
        output_padding, self.groups, self.dilation)

ERROR MESSAGE:
runfile('C:/Users/Priyanka/.spyder-py3/Capsule-Forensics-v2-master/test_binary_ffpp.py', wdir='C:/Users/Priyanka/.spyder-py3/Capsule-Forensics-v2-master')
Reloaded modules: model_big
Namespace(batchSize=32, dataset='databases/faceforensicspp', gpu_id=0, id=21, imageSize=300, outf='checkpoints/binary_faceforensicspp', random=False, test_set='test', workers=0)
Traceback (most recent call last):

File "C:\Users\Admin.spyder-py3\Capsule-Forensics-v2-master\test_binary_ffpp.py", line 63, in
vgg_ext(opt.gpu_id)

File "C:\Users\Admin\Miniconda3\lib\site-packages\torch\nn\modules\module.py", line 493, in call
result = self.forward(*input, **kwargs)

File "C:\Users\Admin.spyder-py3\Capsule-Forensics-v2-master\model_big.py", line 53, in forward
return self.vgg_1(input)

File "C:\Users\Admin\Miniconda3\lib\site-packages\torch\nn\modules\module.py", line 493, in call
result = self.forward(*input, **kwargs)

File "C:\Users\Admin\Miniconda3\lib\site-packages\torch\nn\modules\container.py", line 92, in forward
input = module(input)

File "C:\Users\Admin\Miniconda3\lib\site-packages\torch\nn\modules\module.py", line 493, in call
result = self.forward(*input, **kwargs)

File "C:\Users\Admin\Miniconda3\lib\site-packages\torch\nn\modules\conv.py", line 338, in forward
self.padding, self.dilation, self.groups)

TypeError: conv2d(): argument 'input' (position 1) must be Tensor, not int

Can you please help?

[honghuy127]

I think this is a bug of PyTorch. Could you check the following link?
'ConvTranspose1d' object has no attribute 'padding_mode'