On the main page, you are sharing the generator and encoder for the Model trained on the FFHQ dataset. Can you also share the pre-trained discriminator model?
I tried to convert the tensorflow model (found here: https://github.com/genforce/idinvert) using the stylegan_converter.py from the genforce repository, but I am getting a lot of problems, especially with the tensorflow.contrib.nccl module.
On the main page, you are sharing the generator and encoder for the Model trained on the FFHQ dataset. Can you also share the pre-trained discriminator model? I tried to convert the tensorflow model (found here: https://github.com/genforce/idinvert) using the stylegan_converter.py from the genforce repository, but I am getting a lot of problems, especially with the tensorflow.contrib.nccl module.