[p1esk]

It seems to be empty: https://github.com/lucidrains/DALLE2-pytorch/blob/main/dalle...

[ALittleLight]

If this were a complete project what kind of compute resources would one need to run DALL-E 2? (In inference or train)

[teruakohatu → ALittleLight]

Google Cloud TPUs would be a good start.

[Jack000]

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.

[nullc]

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.

[cfcf14 → nullc]

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).