I'm interested in what you talented people can produce in 1.5 (not SDXL) with NO LoRA, textual inversion or other method such as Lycoris.
NO Negative prompts.
NO img2img!
Basically, only positive prompts can drive the output!
Can do image resize (so generate at 512x768 then img2img at 1024x1536 for example). Any subject, any checkpoint. No NSFW pls, try to be as artistic as possible!!