search

Zhaoyi-Yan / Shift-Net_pytorch

Pytorch implementation of Shift-Net: Image Inpainting via Deep Feature Rearrangement (ECCV, 2018)
http://openaccess.thecvf.com/content_ECCV_2018/papers/Zhaoyi_Yan_Shift-Net_Image_Inpainting_ECCV_2018_paper.pdf
MIT License
363 stars 83 forks source link

Warnings flood the screen #103

Closed muxgt closed 4 years ago

muxgt commented 4 years ago

Hi @Zhaoyi-Yan! I'm training your model in Google Colab. When starting it floods the screen with user warnings during all training. Like this: 1 It is overloading the memory. How can I remove it? Or at least shut it down.

Zhaoyi-Yan commented 4 years ago

It is because that you use pytorch>=1.2, while the guys made ByteTensor deprecated. However, when I make this change, then the required version of this repo will be >=1.2.

Zhaoyi-Yan commented 4 years ago

I have push a commit to solve this, please check!

muxgt commented 4 years ago

Thanks a lot for such a speed! New commit gave error at line 160 in (1 - self.mask_global) saying it doesn't support subtraction in bool so I changed this to ~self.mask_global in line 160 and 177 and now it's working. Thanks again!