>>77496129I haven't looked at her art tag, but I've only seen 5 proper art pieces of the new design posted here and a couple of memes with art of her. Although I guess her already having two different 3D models with the same design also would probably help considering you could add a few screencaps of different angles and the like to the dataset
If you have a decent enough amount of images (like 40 or 50) of her that are different enough, you could probably try zipping it up and requesting someone to run it and try making a lora out of it
I think another tactic that lora makers tend to use when there's a low amount of art is that they just repeat the same few images and make an extremely rough lora that just barely works enough for them to generate one or two more "unique" pieces of art, add that back into the dataset, run it again and do that again an again until they get enough art for an actual functional one. But obviously this takes time and is kind of annoying to do so you likely won't convince a random anon to do it unless they like Tanya to begin with
