janivanecky / artistic-style Goto Github PK

How much time did it take to generate one image? Is it faster than https://github.com/OlavHN/fast-neural-style ?

The code is wonderful and I'm using python 2.7 to run the code.
It outputs the error message: File "/home/jyc/2thPaper_TensorFlow/Artistic-Style-master/vgg.py", line 20, in __init__self.data_dict = np.load(vgg19_npy_path, encoding='latin1').item(). TypeError: load() got an unexpected keyword argument 'encoding' at first. I guess it's due to the python version I used after I refer to http://docs.scipy.org/doc/numpy/reference/generated/numpy.load.html. The code may be developed under python 3.x. So I remove "encoding='latin1'" and it works.

But then I get another error message: "File "style.py", line 111, in input_image = tf.Variable(np.expand_dims(images[0,:,:,:])) TypeError: expand_dims() takes exactly 2 arguments (1 given)"
I wonder if someone has the same error message.

Thanks!

When I opened the link https://dl.dropboxusercontent.com/u/50333326/vgg19.npy to download the "npy file", there was an error message "This account's links are generating too much traffic and have been temporarily disabled!".
Is there a new link available? Thanks!