hello hello ! finally I found a way for generate perfect anime2real workflow that work perfect for every one
no need to talk much but here some tips :
lora models strength better to not be more than 0.8 and additional style model types better to be 0.5
wd14 is fine but you can test for better result
don't try to add another sampler, I already tested it and not worth extra credit cost.
this model can used as img2anime or other styler by changing the models and prompts
I already have 2 AI tool with this workflow : https://tensor.art/template/802619630001757340
(can use it for nsfw)
that is it , and at the end DON'T try to generate image with more than 1400px resolution!!!(it's gonna give you a non sense pics on output 1200px is best)
