Alright, I'm finally happy enough with this LoRA, great results so far so I hope everyone enjoys it aswell.
Recommended:
Strength: 0.9, lower for more flexibility. I'd recommend using a lower cfg for more atmospheric results and higher cfg for more sharp lines. I usually do 4.5 or as low as 2.5.
Some trigger tags also help, such as: wavy_hair, chromatic_aberration, film_grain, halftone, etching, retro_artstyle, 1990s_(style), low contrast, anime coloring, deep shadows
Depending on what you use, it can help and even give you different results improving (and sometimes ruining) your end result, I'm still experimenting as to which extent I can morph the style with prompting, but yeah, that's it.
Thank you again for using my LoRA, you can support me through either donating Buzz, or you can use one of the links provided beneath my profile.
First version, released. This new version is focused more on the "cute 90s aesthetic" retro style, more portrait cute girls, I tried to keep the dataset consistent but it feels like the LoRA is fighting against my quality prompts, so, simpler prompts reinforcing the aesthetic is better than putting 'ultra-HD high definition masterwork cinema' in the prompt, as far as I can tell.
Go ahead and upload yours!
Your query returned no results – please try removing some filters or trying a different term.