5,323
edits
Line 8: | Line 8: | ||
The main difference between stable diffusion and other diffusion models is that the diffusion operations happens in a low-resolution latent space. For a 512x512 image, the latent may only be 64x64, a factor of 8 times smaller. This significantly reduces the compute resources necessary. | The main difference between stable diffusion and other diffusion models is that the diffusion operations happens in a low-resolution latent space. For a 512x512 image, the latent may only be 64x64, a factor of 8 times smaller. This significantly reduces the compute resources necessary. | ||
===Architecture=== | |||
See [https://nn.labml.ai/diffusion/stable_diffusion/model/unet.html U-Net for Stable Diffusion] and [https://nn.labml.ai/diffusion/stable_diffusion/model/unet_attention.html Transformer for Stable Diffusion U-Net] | |||
===1.x=== | ===1.x=== |