XYkong
XYkong
I print he vars, but only "G_synthesis/ToRGB_lod0(/1/2/3/4/5/6)/weight" was found, instead of "G_synthesis/ToRGB/weight". Do i use the wrong weight found on https://nvlabs-fi-cdn.nvidia.com/stylegan2/networks/ ?
It has been solved by using weight files with the suffix f !
> @kongxiuxiu have you got any 256*256 model for pytorch? Yes, and now I have converted 256*256 model for horse, cat and church
> > > @kongxiuxiu have you got any 256*256 model for pytorch? > > > > > > Yes, and now I have converted 256*256 model for horse, cat and...
> > > @kongxiuxiu have you got any 256*256 model for pytorch? > > > > > > Yes, and now I have converted 256*256 model for horse, cat and...
> @kongxiuxiu do they work well? the pics seem weird Most of them looks OK. It works much better than the model trained on my own.
import cv2 import matplotlib.pyplot as plt import matplotlib.image as mpimg import matplotlib as mpl from PIL import Image import numpy as np imgPath1="1video_output/dodo3/stylized_24_stylized_the_resevoir_at_poitiers.jpg" imgPath2="1video_output/dodo3/stylized_25_stylized_the_resevoir_at_poitiers.jpg" import numpy as np img1 =...
We have reviewed the code and you can solve the problem by updating "net.py".
We have revised the code and you can solve the problem by updating "net.py".
It seems you should install "torch" before installing apex.