Trained on NAI.
Recommended weight : 0.9
Trigger words : ariadoa, short hair, red hair, red eyes
Outfit : hair ornament, blue dress, long sleeves, cleavage, large breasts, choker, halterneck, criss-cross halter
If you like the LoRA, reviews are greatly appreciated.
Description
FAQ
Comments (5)
Thanks for the amazing LoRA! I love how flexible it is! Mind sharing how many images you used for the dataset? and did you do any anime screencaps and extra manual tagging?
Hey, thank you for the feedback!
I used 51 images for the dataset, most of them from the anime except for a few. As for the tagging, I use the built-in automatic tagger in the webui, then manually go through the text files to make changes if needed.
@GenCV Thanks for the reply! So you manually looked for screencaps in the anime? 0.0 how tedious was it? did you just screencap every scene with her in it?
@cleopatracleo
I use a website called fancaps, it doesn't contain every single frame of the episodes, but it's usually enough to get a large enough dataset if you go through the entire series.
It can take some time if you count the cropping you have to do afterwards, but it's nothing too crazy. I had to go through episodes myself for other LoRAs, since they weren't on the website, that's when it gets really tedious.
@GenCV Interesting, never knew about the fancaps thing.. well thanks for all the insight!
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.