search

zhanghang1989 / PyTorch-Multi-Style-Transfer

Neural Style and MSG-Net
http://hangzh.com/PyTorch-Style-Transfer/
MIT License
976 stars 207 forks source link

Doesn't work without CUDA: RuntimeError: Input type (CUDAFloatTensor) and weight type (CPUFloatTensor) should be the same #11

Open xaionaro opened 6 years ago

xaionaro commented 6 years ago 
danusya@werewolf:~/PyTorch-Multi-Style-Transfer/experiments$ python main.py train --dataset ~/dataset/    --vgg-model-dir caleido_vgg --save-model-dir caleido_model --epochs 2 --cuda 0
/usr/local/lib/python2.7/dist-packages/torchvision/transforms/transforms.py:156: UserWarning: The use of the transforms.Scale transform is deprecated, please use transforms.Resize instead.
  "please use transforms.Resize instead.")
Net(
  (gram): GramMatrix(
  )
  (model1): Sequential(
    (0): ConvLayer(
      (reflection_pad): ReflectionPad2d((3, 3, 3, 3))
      (conv2d): Conv2d(3, 64, kernel_size=(7, 7), stride=(1, 1))
    )
    (1): InstanceNorm2d(64, eps=1e-05, momentum=0.1, affine=False)
    (2): ReLU(inplace)
    (3): Bottleneck(
      (residual_layer): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2))
      (conv_block): Sequential(
        (0): InstanceNorm2d(64, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(64, 32, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(32, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): ConvLayer(
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(32, 32, kernel_size=(3, 3), stride=(2, 2))
        )
        (6): InstanceNorm2d(32, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(32, 128, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (4): Bottleneck(
      (residual_layer): Conv2d(128, 512, kernel_size=(1, 1), stride=(2, 2))
      (conv_block): Sequential(
        (0): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): ConvLayer(
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2))
        )
        (6): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1))
      )
    )
  )
  (ins): Inspiration(N x 512)
  (model): Sequential(
    (0): Sequential(
      (0): ConvLayer(
        (reflection_pad): ReflectionPad2d((3, 3, 3, 3))
        (conv2d): Conv2d(3, 64, kernel_size=(7, 7), stride=(1, 1))
      )
      (1): InstanceNorm2d(64, eps=1e-05, momentum=0.1, affine=False)
      (2): ReLU(inplace)
      (3): Bottleneck(
        (residual_layer): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2))
        (conv_block): Sequential(
          (0): InstanceNorm2d(64, eps=1e-05, momentum=0.1, affine=False)
          (1): ReLU(inplace)
          (2): Conv2d(64, 32, kernel_size=(1, 1), stride=(1, 1))
          (3): InstanceNorm2d(32, eps=1e-05, momentum=0.1, affine=False)
          (4): ReLU(inplace)
          (5): ConvLayer(
            (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
            (conv2d): Conv2d(32, 32, kernel_size=(3, 3), stride=(2, 2))
          )
          (6): InstanceNorm2d(32, eps=1e-05, momentum=0.1, affine=False)
          (7): ReLU(inplace)
          (8): Conv2d(32, 128, kernel_size=(1, 1), stride=(1, 1))
        )
      )
      (4): Bottleneck(
        (residual_layer): Conv2d(128, 512, kernel_size=(1, 1), stride=(2, 2))
        (conv_block): Sequential(
          (0): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
          (1): ReLU(inplace)
          (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
          (3): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
          (4): ReLU(inplace)
          (5): ConvLayer(
            (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
            (conv2d): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2))
          )
          (6): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
          (7): ReLU(inplace)
          (8): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1))
        )
      )
    )
    (1): Inspiration(N x 512)
    (2): Bottleneck(
      (conv_block): Sequential(
        (0): InstanceNorm2d(512, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): ConvLayer(
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1))
        )
        (6): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (3): Bottleneck(
      (conv_block): Sequential(
        (0): InstanceNorm2d(512, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): ConvLayer(
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1))
        )
        (6): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (4): Bottleneck(
      (conv_block): Sequential(
        (0): InstanceNorm2d(512, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): ConvLayer(
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1))
        )
        (6): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (5): Bottleneck(
      (conv_block): Sequential(
        (0): InstanceNorm2d(512, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): ConvLayer(
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1))
        )
        (6): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (6): Bottleneck(
      (conv_block): Sequential(
        (0): InstanceNorm2d(512, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): ConvLayer(
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1))
        )
        (6): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (7): Bottleneck(
      (conv_block): Sequential(
        (0): InstanceNorm2d(512, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): ConvLayer(
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1))
        )
        (6): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (8): UpBottleneck(
      (residual_layer): UpsampleConvLayer(
        (upsample_layer): Upsample(scale_factor=2, mode=nearest)
        (conv2d): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1))
      )
      (conv_block): Sequential(
        (0): InstanceNorm2d(512, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(512, 32, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(32, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): UpsampleConvLayer(
          (upsample_layer): Upsample(scale_factor=2, mode=nearest)
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1))
        )
        (6): InstanceNorm2d(32, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(32, 128, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (9): UpBottleneck(
        (upsample_layer): Upsample(scale_factor=2, mode=nearest)
        (conv2d): Conv2d(128, 64, kernel_size=(1, 1), stride=(1, 1))
      )
      (conv_block): Sequential(
        (0): InstanceNorm2d(128, eps=1e-05, momentum=0.1, affine=False)
        (1): ReLU(inplace)
        (2): Conv2d(128, 16, kernel_size=(1, 1), stride=(1, 1))
        (3): InstanceNorm2d(16, eps=1e-05, momentum=0.1, affine=False)
        (4): ReLU(inplace)
        (5): UpsampleConvLayer(
          (upsample_layer): Upsample(scale_factor=2, mode=nearest)
          (reflection_pad): ReflectionPad2d((1, 1, 1, 1))
          (conv2d): Conv2d(16, 16, kernel_size=(3, 3), stride=(1, 1))
        )
        (6): InstanceNorm2d(16, eps=1e-05, momentum=0.1, affine=False)
        (7): ReLU(inplace)
        (8): Conv2d(16, 64, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (10): InstanceNorm2d(64, eps=1e-05, momentum=0.1, affine=False)
    (11): ReLU(inplace)
    (12): ConvLayer(
      (reflection_pad): ReflectionPad2d((3, 3, 3, 3))
      (conv2d): Conv2d(64, 3, kernel_size=(7, 7), stride=(1, 1))
    )
  )
)
Traceback (most recent call last):
  File "main.py", line 287, in <module>
    main()
  File "main.py", line 40, in main
    train(args)
  File "main.py", line 159, in train
    style_model.setTarget(style_v)
  File "/home/danusya/PyTorch-Multi-Style-Transfer/experiments/net.py", line 293, in setTarget
    F = self.model1(Xs)
  File "/usr/local/lib/python2.7/dist-packages/torch/nn/modules/module.py", line 357, in __call__
    result = self.forward(*input, **kwargs)
  File "/usr/local/lib/python2.7/dist-packages/torch/nn/modules/container.py", line 67, in forward
    input = module(input)
  File "/usr/local/lib/python2.7/dist-packages/torch/nn/modules/module.py", line 357, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/danusya/PyTorch-Multi-Style-Transfer/experiments/net.py", line 153, in forward
    out = self.conv2d(out)
  File "/usr/local/lib/python2.7/dist-packages/torch/nn/modules/module.py", line 357, in __call__
    result = self.forward(*input, **kwargs)
  File "/usr/local/lib/python2.7/dist-packages/torch/nn/modules/conv.py", line 282, in forward
    self.padding, self.dilation, self.groups)
  File "/usr/local/lib/python2.7/dist-packages/torch/nn/functional.py", line 90, in conv2d
    return f(input, weight, bias)
RuntimeError: Input type (CUDAFloatTensor) and weight type (CPUFloatTensor) should be the same
LoicEm commented 6 years ago

There seem to be a problem where some of the data is loaded on the GPU even with CUDA disabled. Did you test it successfully with CUDA ?

liangbh6 commented 6 years ago

I got the same problem with pytorch=0.3.1 too. Have you found the solution?