v1.0
This lora can help you to finetune your images in connection with 20 nationalities. The model was trained on handpicked UHD images of female and male models of the given nationality.
Here are the nationalities what the model was trained on:
African
African-American
American
Arab
Asian-American
Brazilian
Chinese
English
Finnish
French
German
Indian
Israeli
Italian
Norwegian
Spanish
Swedish
Taiwanese
Turkish
Ukrainian
Taiwanese is not complete yet but the rest of them should be able to adjust the images.
Strength:
SDXL: it seems like it works well around 1.00 up to 1.30-1.50
PONY:
the effect is very strong here, 0.05 might create visible changes already if you want to keep the anime look - the higher you go, the more realistic the images will be
if you fancy a more anime look then go in the negative direction
See examples in the gallery for all these.
Have fun exploring it and share your results!
Description
FAQ
Comments (10)
Korean?
hello, there are around 200 countries in the world, so there is a lot left out :-) at the moment this is more of a test to see if this works and whether people use it. we will see the rest later :) but thanks for the suggestion anyways!
@maxxxx_civitai Thank you so much. I can't wait to see how this grows. Good luck!
Love this LORA, great Ukrainian, English and American Asian models. Love your work
thanks, have fun with it :)
thank you for making!
will version 2 include Pacific Islander people, like Filipinos, etc.? :)
hi,
glad you like it:)
well, first I needed to see if and how it works. How it can be expanded will be phase two. But there are some 200 countries and this inlcudes just 20 nationalities, so, there is room for improvement :)))
@maxxxx_civitai no pressure. :) thank you very much for doing this <3
I can't remember which ones (realistic vision I think is one) but a few of the sd1.5 checkpoints recognize nationalities, even to the point that if I use Congolese they're beefy and with Eritrean they're wiry; you've probably already noticed this. I haven't tried this Lora yet but will do so; I'm so tired of that "Pony guy" face in Pony derived checkpoints. So a big thanks.
hey,
did not know:) don't expect this level of dictinction though here, this is not a checkpoint :D but I hope you will like it :)