generator model keys doesn't match

[yeong95]

I trained stylegan generator from genforce (https://github.com/genforce/genforce/tree/2ad04974cfaeba20b93c806531f987f06cc5c328)

But when loading weights from trained model the keys doesn't match each other.

How do i fix this?

[zhujiapeng]

The stylegan generator of genforce is refactored. The whole training pipeline can be found here, this pytorch repo only supports inference.

[yeong95]

Thank you for reply.

And one more question: Is the inversion process in genforce repo the same as the inversion process in this repo?

[zhujiapeng]

No, the genforce supports training the encoder in this repo.