git-re-basin VS VToonify

Implementation: https://github.com/samuela/git-re-basin

Wouldn't "applying the permutation" simply swap all the parameters in a model so they match on both models? For example, in https://github.com/samuela/git-re-basin/blob/main/src/cifar10_vgg_weight_matching.py, on line 184 they apply the permutation, and on line 192 they lerp from model A's params to the permuted model B's params. This lerp is basically a weighted sum merge, isn't it? At a lerp of 0.5, it would be somewhere in between model A and the permuted model B.

Definitely interesting, but you might get something useful out of https://github.com/samuela/git-re-basin ?

Git Re-Basin: Merging Models modulo Permutation Symmetries https://github.com/samuela/git-re-basin https://arxiv.org/abs/2209.04836v1

this is a VToonify you can chek this here https://github.com/williamyang1991/VToonify

Here is the software repository

Vtoonify is a machine learning model (GAN) created that unlike stable diffusion always reproduces the same output given the same input. For example my character Midas went in looking like the right, can came out looking like the left.