Live data from Hacker News

Untitled topic

news.ycombinator.com

1–7 of 7 posts

Re: undefined

#4
I hope they find latent diffusion works for this, without it this will probably be too expensive for private parties to train on big collections.

Re: undefined

#5
post #4

I hope they find latent diffusion works for this, without it this will probably be too expensive for private parties to train on big collections.

Isn't the VQ-VAE/dVAE generator approach in the DALL-E models quite a bit cheaper computationally than latent diffusion models?

My understanding was that diffusion models were quite a bit more expensive, but yielded richer latent distributions and better images (for some definition of better).

Re: undefined

#7
hm.. I wonder which clip model they'll use. A big part of what makes DALLE-2 so good is the unreleased huge clip model. To train the diffusion prior they may need to first replicate this clip model.