Stable Diffusion 2.0 Release

Re:Stable Diffusion 2.0 Release by hackater on Wednesday November 23, 2022 @11:47PM (#33727896)

Can you go into a bit more detail? What architecture did you use? Is the month training time really just training with mini batches with a constant learning rate? Or are these many failed attempts until you trained a successful model for a few days in the end?

I particularly interested in the image generation part (the DDPM/SGM)

[ Parent | on HN ]

Re:Stable Diffusion 2.0 Release by Jack000 on Thursday November 24, 2022 @12:02AM (#33727949)

Yeah I did have a few false starts. Total time is more like 3 months vs 1 month for the final model. For small scale training I found it’s necessary to use a long lr warmup period, followed by constant lr.

There’s code on my GitHub (glid3)

edit: The architecture is identical to SD except I trained on 256px images with cosine noise schedule instead of linear. Using the cosine schedule makes the unet converge faster but can overfit if overtrained.

[ Parent | on HN ]

Re:Stable Diffusion 2.0 Release by rrobukef on Thursday November 24, 2022 @04:38AM (#33729390)

I keep wondering if using not only statistical noise but also deformations would help with the generation of deformable things - say human hands.

[ Parent | on HN ]

Re:Stable Diffusion 2.0 Release by dmingod666 on Thursday November 24, 2022 @06:39AM (#33730199)

F222 does a little more coherent anatomy..not surprising given its background

[ Parent | on HN ]