IMPORTANT: This LoRA is trained as a V-Pred model and is fully compatible with other V-Pred models. For the best results, make sure to use it with a V-Pred checkpoint.
Trigger words:
octaviaHB, beak, colored sclera, pink sclera, white pupils, black hair, long hair
Outfit:
beanie, pink choker, black cardigan, long sleeves, hat, black leggings
Tested and used with WAI-SHUFFLE-NOOB:
https://civarchive.com/models/989367?modelVersionId=1202076
For on-site generations, use my Tensor Art page: https://tensor.art/u/830771749061779884
Join my Patreon for full access:
https://www.patreon.com/Waifu_Chef
Follow me on Twitter:
Check out my Pixiv gallery:
🎨 https://www.pixiv.net/en/users/111915639
Order a commission here!
https://ko-fi.com/c/38aa4a973d
Support my work by buying me a ko-fi:
Description
FAQ
Comments (4)
This lora (and many from Waifu_Chef) are great, and I genuinely recommend them. I just wished they had a more art-style varied dataset. The show's style is extremely overcooked on all LoRAs I've tried. I get 0 creativity unless I use 0.25 weights or less and by then I am counting on the base model having a faint idea of the character to start with.
I tried different models to make sure and it happens on all of them:
YiffyMix, Augenstein, YiffinHell, Whatastic, NAI, etc.
This hurts me because for Octavia, this is the only LoRA that doesn't give her a big ass beak. Same with other characters.
If I am missing something, please let me know.
Thank you, I appreciate the feedback! My priority is maximizing character accuracy, so the datasets are built exclusively from official screencaps rather than fan art. I understand that this can make the style very strong, but it helps keep the identity consistent.
@Waifu_Chef Keep cooking.
@GhostyWolff Oh I will! More on the way 🍳🔥