About

I'm Olga, debru online. I got into AI images not long ago: first I played with online services, then I wanted to understand what was going on inside, so I switched to generating locally.

My setup

• An RTX 4060 Ti with 16 GB — plenty for SDXL; a 1024×1024 image at 30 steps takes a few seconds.
• Python and the diffusers library; sometimes ComfyUI for node experiments.
• Models: Stable Diffusion XL base 1.0 and SDXL Turbo.

About this site

Themed series and notes on technique: what steps, CFG, seed and negative prompts actually do. Every image comes with its full settings, so you can reproduce it (give or take library versions and hardware).

All images are generated by me. You can use them under CC BY 4.0 — with a link back to this site.