Is a separate model trained per benchmark?

[RylanSchaeffer]

trafficstars

I have a quick question. In the paper, is a separate model trained per-dataset? Or is one model trained for all datasets?

[rromb]

For the different conditioning tasks (semantic synthesis, depth-to-image etc) we train different transformer models. The VQGAN on ImageNet should be fairly general and we re-use it across some tasks, but for specialized tasks such as face synthesis we train dedicated VQGANs. Hope this helps!