Same, mainly due to having a intel i5 computer that I bought for $20 . I know you can use xl lora models along sd1.5 - 2.0 . I'm not sure about, 3.0 since, I can't test that right now. With comfy ui and stable diffussion gui by nmkd. The lora xl models still work as shader models from what I could test. I'm not sure how well the lora models handle on S.D. by nmkd since, I don't have a nvidia gui. So, I don't know how well the models handle on local hardware ^.^T

Thanks, I tried to be more picky with the models. For Werethrope's art style. An I just finished adding new models for his oc Werethrope, and ok ki-ru. I plan on training models on the other characters. But, I'll have to rip the models from the game boof vr. To render different angles to make those models. Since, 2 or 3 characters from boof vr doing have much good refference art.

I actually have other artist models I wanted to do. But, I'm trying to make sure some of the styles don't give off the ugly 3d look some stable diffusion models give off. I have hundreds of images of photos I took. To help train a more atmospheric furry model. But, I need alot of morning, sunset, and night time pics. So, I can merge those images with some furry art. To hopefully make a new model that can make anthro characters look good while not being uncanny. So that model might be a experimental thing .