no attribute PerceptualLoss about real-world-sr

[TouqeerAhmad]

Are you by any chance using an older version of LPIPS. The following excerpt from LPIPS introduction points there are two versions:

"If you use LPIPS in your publication, please specify which version you are using. The current version is 0.1. You can set version='0.0' for the initial release."

[TouqeerAhmad]

Never mind about this issue. I was able to get rid of this by importing models and replacing ps with models when initializing the loss network. Specifically, following two lines should address this if anyone else gets the same error:

'from PerceptualSimilarity import models'
'self.loss_network = models.PerceptualLoss(use_gpu=torch.cuda.is_available())'

[hcleung3325]

Never mind about this issue. I was able to get rid of this by importing models and replacing ps with models when initializing the loss network. Specifically, following two lines should address this if anyone else gets the same error:

'from PerceptualSimilarity import models'
'self.loss_network = models.PerceptualLoss(use_gpu=torch.cuda.is_available())'

Hi Ahmad,
Thanks for your suggestion.
I download the original version, it can import now.
However, there is another problem.

Saving summary into directory runs/save_model/0/
[1/300]: 0%| | 0/50 [00:15<?, ?it/s]
Traceback (most recent call last):
File "train.py", line 186, in
g_loss.backward()
File "/home/root/miniconda/envs/py37/lib/python3.7/site-packages/torch/tensor.py", line 185, in backward
torch.autograd.backward(self, gradient, retain_graph, create_graph)
File "/home/root/miniconda/envs/py37/lib/python3.7/site-packages/torch/autograd/init.py", line 127, in backward
allow_unreachable=True) # allow_unreachable flag
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [1, 256, 1, 1]] is at version 2; expected version 1 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).

Have u experienced this before?

[jfun9494]

@hcleung3325 I remember solving this by using torch==1.1.0

[wxslby]

self.loss_network

Never mind about this issue. I was able to get rid of this by importing models and replacing ps with models when initializing the loss network. Specifically, following two lines should address this if anyone else gets the same error:

'from PerceptualSimilarity import models' 'self.loss_network = models.PerceptualLoss(use_gpu=torch.cuda.is_available())'

Never mind about this issue. I was able to get rid of this by importing models and replacing ps with models when initializing the loss network. Specifically, following two lines should address this if anyone else gets the same error:

'from PerceptualSimilarity import models' 'self.loss_network = models.PerceptualLoss(use_gpu=torch.cuda.is_available())'

Sorry, I can't find "from PerceptualSimilarity import models".Where is the "models"? Can you share the code ?

[mrgreen3325]

PerceptualSimilarity

You can download the PerceptualSimilarity package.
Or I email you the package, what is your email?

[wxslby]

Thanks for your reply. My email is [email protected]. Please!

[cgysjy]

The downloaded perceptualsimilarity works, but still reports an error:
AttributeError: module 'PerceptualSimilarity' has no attribute 'PerceptualLoss'.
The current lpips minimum version 1.0 cannot be reduced to 0.0.
I did not find models for the above solution.
Is there any other solution? Thank you!