There're two versions of Hioshiru's artstyle Lora model. One is trained on FurryVixen base model, another is on Lucid's Mix. You can find both base model on Civitai.
I trained it on 301 images I downloaded from e621 and tagged it with wd-tagger.
Tips: you can use it together with other characters' Lora and it works well. No trigger words are needed. Hope you enjoy it.
Some images examples are not Upscaled.
It is recommoned that you use anything's VAE or some other VAE if you want to get high saturation results.
After I install Locon Support extension in WebUI, the result seems to get very bad. I don't know the reason. Weird. I don't recommond to install that Support. If it reports 'Failed to match keys when loading Lora', just ignore it.
Description
Trained on FurryVixen base model.
FAQ
Comments (3)
Small detail.. some of your images for each version are mixed. Might be better to seperate them for each version so they are showing realistic expectations for each.
AWESOME WORK BTW!! keep it up
Where can one get a copy of LucidFurry?
its just a rename of lucid's mix your good with the base model.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.