Try using the original adversarial losses for the 2D LDM tutorial

[Warvito]

In the 3D examples, we had problems with the original LDM's adversarial losses. This does not mean it will not work well for the 2D scenarios.

[QuantPrincess]

@Warvito Can you elaborate more on this? I'm having trouble getting good results for the diffusion part of the pipeline for 3d LDM. My autoencoding model of choice (vqgan) seems to be reasonably converged, but when using the template diffusion model in the 3d LDM tutorial with my autoencoding model, diffusion results aren't quite able to capture the necessary structure and detail of my objects. Did you run into this issue ever? I have an image sice of 64^3.

[blofn]

I meet the same problem，3dvqgan’s results are good，but LDM can't generate right latent vector.