Why it matters
SD family dominates open image generation. Understanding shapes vision generation.
Advertisement
The architecture
Text encoder: CLIP or T5.
UNet: denoises latents.
VAE: latent to pixel.
Advertisement
How it works end to end
Training: predict noise added to latent.
Inference: 20-50 denoising steps.
Schedulers: DDIM, DPM++.
ControlNet: conditioning.