Encoder / Decoder Configuration?
#11
by
gturk1
- opened
I am reading through the Latent Diffusion Model paper, and they describe two different kinds of encoder /decoder that they tested in their model, VQ and KL. Does anyone know which model (VQ or KL) was used to create the weights that we are using? Thanks!
Somebody can correct me if I'm wrong, but I think we are using KL.