Stability-AI / stablediffusion

High-Resolution Image Synthesis with Latent Diffusion Models
MIT License
38.83k stars 5.01k forks source link

What is the encoding/color model for the SDXL latents? #335

Open TimothyAlexisVass opened 11 months ago

TimothyAlexisVass commented 11 months ago

I'm kinda groping in the dark. So I would be grateful if someone could tell me.

Have gotten this far by treating the channels as Luma, ChromaCyan, ChromaGreen and Key: image

So my question is simply, what are the 4 channels in the latents tensor?