abhiskk/fast-neural-style

Issues

• trained model not working

  #19 opened 7 years ago by ksesalebangim
  1

• About adding Tanh() at the end of the model

  #12 opened 8 years ago by zhanghang1989
  5

• RuntimeError: The expanded size of the tensor (700) must match the existing size (32) at non-singleton dimension 2

  #30 opened 7 years ago by dovanchan
  1

• Python error: <stdin> is a directory, cannot continue

  #29 opened 7 years ago by dovanchan
  1

• Could you upload vgg16.weight file?

  #28 opened 7 years ago by jmhong-simulation
  1

• Query Regarding Transforms used for input image and style image

  #26 opened 7 years ago by ssundar6087
  5

• download_styling_models.sh error

  #27 opened 7 years ago by mqchen1993
  0

• train doesn't run: ValueError: not enough values to unpack (expected 3, got 2)

  #25 opened 7 years ago by xaionaro
  5

• content loss calculation is wrong?

  #24 opened 7 years ago by RyanCV
  1

• some question in utils.py

  #23 opened 7 years ago by poptree
  1

• download_styling_models.sh error

  #22 opened 7 years ago by bliunlpr
  1

• difference with jcjohnson

  #20 opened 7 years ago by onlytailei
  1

• Time for training a new style

  #17 opened 7 years ago by nitish116
  3

• Problem with training

  #21 opened 7 years ago by victorkitov
  0

• How to set output_image_size when Eval

  #14 opened 7 years ago by dovanchan
  18

• Runtime error while trying out the toy example

  #15 opened 7 years ago by whiletruelearn
  2

• Check usage of 'subtract_imagenet_mean', with new version of pytorch inplace subtraction might be different.

  #11 opened 8 years ago by abhiskk
  1

• stylize using GPU

  #9 opened 8 years ago by jayavanth
  19

• Use median filter

  #8 opened 8 years ago by abhiskk
  0

• Use Upsampling instead of ConvTranspose2d

  #6 opened 8 years ago by abhiskk
  0

• Use Reflection Padding

  #7 opened 8 years ago by abhiskk
  0

• Use instance normalization

  #2 opened 8 years ago by abhiskk
  0

• Add code for stylizing images, use model.eval() when you are stylizing.

  #3 opened 8 years ago by abhiskk
  0

• In the default configuration don't resize the styling-image when training the network

  #4 opened 8 years ago by abhiskk
  0

• Save CPU version of the trained model

  #5 opened 8 years ago by abhiskk
  0

• The scale of content and style loss just changed from 1e5 (content), 1e3 (style) to 1e2(content), (1e-2) style.

  #1 opened 8 years ago by abhiskk
  0