VAE in Training of each LDM in CoDi

[NakataKoo]

Thank you for the great project.

Am I right in thinking that "in learning each Latent Diffusion Model in CoDi, the Encoder-Decoder part of VAE uses a pre-trained model, so only the U-Net part is the target of training"?

Also, do you have any plans to release the study scripts?

[danielemolino]

I'll to know it too