folder structure (* model that is only for training, not required for inference):
checkpoint
|--encoder.pt % Pixel2style2pixel model
|--faceparsing.pt % Face parsing model
|--stylegan2-ffhq-config-f.pt % * StyleGAN model
|--derections.npy % * Editing vectors
% the followings are VToonify models
|--vtoonify_t_cartoon
|--pretrain.pt % * Pre-trained encoder for Cartoon style
|--vtoonify.pt % VToonify-T model for Cartoon style
|--vtoonify_d_cartoon
|--pretrain.pt % * Pre-trained encoder for Cartoon style
|--exstyle_code.npy % Extrinsic style codes of Cartoon dataset
|--vtoonify_s_d.pt % VToonify-Dsd model for Cartoon style
|--vtoonify_s_d_c.pt % VToonify-Dsdc model for Cartoon style
|--vtoonify_s026_d0.5.pt % VToonify-D model for the 26th Cartoon style with style degree 0.5
... % VToonify-D model for other settings
|--vtoonify_t_caricature
% the same files as in vtoonify_t_cartoon
|--vtoonify_d_caricature
% the same files as in vtoonify_d_cartoon
...
% the followings are pre-trained Toonify and DualStyleGAN models
|--cartoon
|--generator.pt % * DualStyleGAN model
|--exstyle_code.npy % * Extrinsic style codes of Cartoon dataset
|--refined_exstyle_code.npy % * Refined extrinsic style codes of Cartoon dataset
|--finetune-000600.pt % * Toonify model (fine-tuned StyleGAN on Cartoon dataset)
|--caricature
% the same files as in cartoon
...