About the Toonify Classic (original model)

[Shysto]

@justinpinkney
Hello, first of all thank you for providing such an awesome tool! It is really a great job!

I was wondering if you would agree to share the "original toonify" Pytorch model that has been trained on pix2pixHD (or is it already shared somewhere?)?
As you explained in your blog, the optimization process (used in your Google Colab notebook) for finding the latent of a real person's face is really expensive...

Would it be possible to share this pre-trained model (on 1024x1024 images if you have), before the ONNX exportation (i.e. the .pth file)?

If not, would you mind to share more details about the training of this model on pix2pixHD. so it could be reproducible?
(e.g. is the dataset available? If not, how many pairs of images FFHQ/Blended-model generated did you use for training? How many iterations? How much time did it take?...).

Thank you very much!