hello,thanks for your codes, can you shre your trained model to test my own content images? thank you

hell,The pre-trained model can't be download.Can you give a model to me? Thank you.

Hi!

I am really interested in this concept, and I think you are really somebody, I mean expert🕶️.

As you can see some small textures in the dog, and I think bigger textures are greater(look at the style image, cat)

Do you have tried about the scales that affect the final generated images?

Like this ⬆️

Thank you so much! I am trying this as well but it may take longer.

There is an error when the model is being sliced for the encoder.

/tmp/ipykernel_10779/972606360.py in fit(epochs, optimizer, model_state_dir, iters, learning_rate_decay)
7 for (i, (content,style)) in progress_bar(enumerate(zip(content_dl,style_dl), 1)):
8 #style = next(style_iter)
----> 9 loss = model(content, style)
10 loss_list.append(loss.item())
11 #adjust_learning_rate(optim, iters, learning_rate_decay)

~/anaconda3/envs/pytorch_p39/lib/python3.9/site-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
1192 if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
1193 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1194 return forward_call(*input, **kwargs)
1195 # Do not call functions when jit is used
1196 full_backward_hooks, non_full_backward_hooks = [], []

/tmp/ipykernel_10779/887166169.py in forward(self, content_images, style_images, alpha, lam)
27
28 def forward(self, content_images, style_images, alpha=1.0, lam=10):
---> 29 content_features = self.vgg_encoder(content_images, output_last_feature=True)
30 style_features = self.vgg_encoder(style_images, output_last_feature=True)
31 t = adain(content_features, style_features)

~/anaconda3/envs/pytorch_p39/lib/python3.9/site-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
1192 if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
1193 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1194 return forward_call(*input, **kwargs)
1195 # Do not call functions when jit is used
1196 full_backward_hooks, non_full_backward_hooks = [], []

/tmp/ipykernel_10779/2727537595.py in forward(self, images, output_last_feature)
11
12 def forward(self, images, output_last_feature=False):
---> 13 h1 = self.slice1(images)
14 h2 = self.slice2(h1)
15 h3 = self.slice3(h2)

~/anaconda3/envs/pytorch_p39/lib/python3.9/site-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
1192 if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
1193 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1194 return forward_call(*input, **kwargs)
1195 # Do not call functions when jit is used
1196 full_backward_hooks, non_full_backward_hooks = [], []

~/anaconda3/envs/pytorch_p39/lib/python3.9/site-packages/torch/nn/modules/container.py in forward(self, input)
202 def forward(self, input):
203 for module in self:
--> 204 input = module(input)
205 return input
206

~/anaconda3/envs/pytorch_p39/lib/python3.9/site-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
1192 if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
1193 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1194 return forward_call(*input, **kwargs)
1195 # Do not call functions when jit is used
1196 full_backward_hooks, non_full_backward_hooks = [], []

~/anaconda3/envs/pytorch_p39/lib/python3.9/site-packages/torch/nn/modules/conv.py in forward(self, input)
461
462 def forward(self, input: Tensor) -> Tensor:
--> 463 return self._conv_forward(input, self.weight, self.bias)
464
465 class Conv3d(_ConvNd):

~/anaconda3/envs/pytorch_p39/lib/python3.9/site-packages/torch/nn/modules/conv.py in _conv_forward(self, input, weight, bias)
457 weight, bias, self.stride,
458 _pair(0), self.dilation, self.groups)
--> 459 return F.conv2d(input, weight, bias, self.stride,
460 self.padding, self.dilation, self.groups)
461

TypeError: conv2d() received an invalid combination of arguments - got (list, Parameter, Parameter, tuple, tuple, tuple, int), but expected one of:

• (Tensor input, Tensor weight, Tensor bias, tuple of ints stride, tuple of ints padding, tuple of ints dilation, int groups)
  didn't match because some of the arguments have invalid types: (!list!, !Parameter!, !Parameter!, !tuple!, !tuple!, !tuple!, int)
• (Tensor input, Tensor weight, Tensor bias, tuple of ints stride, str padding, tuple of ints dilation, int groups)
  didn't match because some of the arguments have invalid types: (!list!, !Parameter!, !Parameter!, !tuple!, !tuple!, !tuple!, int)

F:\arbitrary\Pytorch_Adain_from_scratch-master\Pytorch_Adain_from_scratch-master>python test.py -c f:\arbitrary\Pytorch_Adain_from_scratch-master\Pytorch_Adain_from_scratch-master\content\tree.jpg -s f:\arbitrary\Pytorch_Adain_from_scratch-master\Pytorch_Adain_from_scratch-master\style\news1.jpg
Traceback (most recent call last):
File "test.py", line 84, in
main()
File "test.py", line 64, in main
res = torch.cat([c_denorm, out_denorm], dim=0)
RuntimeError: invalid argument 0: Sizes of tensors must match except in dimension 0. Got 459 and 456 in dimension 2 at C:\w\1\s\windows\pytorch\aten\src\TH/generic/THTensor.cpp:689

content图片

style图片

How long did you train your AdaIN network and I have another question.

Do you have the problem that the pixel tends gray. Like this.

I trained model for 20 epochs. It seem doesn't learn anything. Can you help me to fix it?

Hi @irasin , I am very interested in this image-to-image translation translation without Cycle Consistency Loss. I am confused about wether this method can be used in unpaired image-to-image? I think maybe can achieve it by changing datasets.py.

Hello , irasin. Thank you very much for your excellent project with pytorch . I am very interested in this project. However , I am puzzled about the denorm operation in training ,
with torch.no_grad():
out = model.generate(content, style)
content = denorm(content, device)
style = denorm(style, device)
out = denorm(out, device)
res = torch.cat([content, style, out], dim=0)
res = res.to('cpu')
save_image(res, f'{image_dir}/{e}epoch{i}_iteration.png', nrow=args.batch_size)
I found that the out is not satisfactory without denorm operation. But I can't understand why use the denorm here ? when we should add this operation? Thank you for your project and looking for your help . Good weekend for you.