Type | |
Stats | 304 1,083 37 |
Reviews | (97) |
Published | Jun 19, 2024 |
Base Model | |
Usage Tips | Clip Skip: 2 Strength: 1 |
Trigger Words | pbj peanutbutter jelly jam |
Hash | AutoV2 1F2C917F8C |
ITS PEANUT BUTTER JELLY TIME! (jam in some parts of the world)
here is a lora for turning things into peanutbutter and jelly
the magic words are
pbj: this is the entire dataset
peanutbutter: notice no spaces, when you have a space it will take more from the model and may be better or worse depending on what youre doing
jelly: this is jelly
jam: this is jam
special note
I was told grape jelly is yucky so the dataset was not trained on it, if you mess around you can probably do it though
the lora isn't perfect, I didn't get a very extensive dataset so it may struggle with realism and may take some supporting tags I've laid out at the end to get your desired outcome.
the lora from my testing didn't seem to work well with realism pony checkpoints or sdxl realism checkpoints, it may be because food items are better trained on those or that the dataset simply doesn't adapt well because of its styles. but feel free to try it out, you might figure out how to do it.
some good positives to use: feral,melting, slime, goo,full_body,no_humans,solo,
some negatives that helped me: collar,jar,source_pony,source_furry