yijunmaverick / cartoongan-test-pytorch-torch Goto Github PK

File "test.py", line 27, in
model.load_state_dict(torch.load(os.path.join(opt.model_path, opt.style + '_net_G_float.pth')))
File "/home/qz/qz/qz/lib/python3.5/site-packages/torch/serialization.py", line 368, in load
return _load(f, map_location, pickle_module)
File "/home/qz/qz/qz/lib/python3.5/site-packages/torch/serialization.py", line 532, in _load
magic_number = pickle_module.load(f)
_pickle.UnpicklingError: invalid load key, '<'.
I run the programme on torch-1.0.1, not torch-0.3, will this affect the outcome?

Hello！brother! I have reproduced the CartoonGan, according to the content of the paper, but the result is not good, some content of the original image have distortion, and the the training is a little shaky, do you have some idea?

How to create data sets with fuzzy edges?

RuntimeError: cuda runtime error (2) : out of memory at /opt/pytorch/pytorch/torch/lib/THC/generic/THCStorage.cu:58

Hi,

Is it possible to produce larger output images? Currently, it seems the outputs are around 450x300. I tried adding a --load_size 1024 option but it returns with " TypeError: can't multiply sequence by non-int of type 'float' " . Would you happen to know how to generate larger images around 1024x1024? Thanks.

What's the License for this github ?
there is no comment for this github .

Thanks a lot for your code. But when I run the torch version code, I got some pure color pictures.

http://vllab1.ucmerced.edu/~yli62/CartoonGAN/pytorch_pth/Hayao_net_G_float.pth
This url shows 404

Hi
I'm interested your paper
So I'm reproducing the CartoonGAN

However, I don't know hyper-paramter in detail.. like weight initialization.. (xavier ? gaussian?)
I want to know hyper-parameter information

1. Optimizer (if Adam, what beta1, beta2` ?)
2. Weight initialization
3. Use l2 regularization (weight decay) ?
4. Data augmentation
5. Normalization function in discriminator (batch norm? layer norm?)
6. Loss function (general gan loss ? or lsgan loss like cyclegan ?)

and can you share the animation dataset used in your paper?

Thank you !

Can you share your tool that transform the torch model to Pytorch model? Thank you very much.

File "video2cartoon.py", line 7, in
from model.Transformer import Transformer
ModuleNotFoundError: No module named 'model'

I found that if not include the RGB -> BGR and BGR -> RGB,the results still normal.

Hi, I was just kindly wondering if I could use an image that is converted to cartoon style using your model in a commercial work, like a background image of a poster or a game? Thanks a lot!

first ,thank your work？ I want to know why use ReflectionPad2d in the test pahse？

@Yijunmaverick
I tried the model on my local i3 CPU and could reproduce the results for the images already given in the code. However when I present some random images from my desktop the results arent good.

Any reason why this might be happening?

Any idea how to fix it?

(base) C:\Users\anton>D:\DFL\Anaconda\CartoonGAN-Pytorch-master\test.py
Traceback (most recent call last):
File "D:\DFL\Anaconda\CartoonGAN-Pytorch-master\test.py", line 25, in
shutil.rmtree(opt.output_dir)
File "D:\DFL\Anaconda\lib\shutil.py", line 500, in rmtree
return _rmtree_unsafe(path, onerror)
File "D:\DFL\Anaconda\lib\shutil.py", line 382, in _rmtree_unsafe
onerror(os.listdir, path, sys.exc_info())
File "D:\DFL\Anaconda\lib\shutil.py", line 380, in _rmtree_unsafe
names = os.listdir(path)
FileNotFoundError: [WinError 3] The system cannot find the path specified: './output_lcy'