The dataset includes works by multiple artists in Blue Archive.
7peach(Including Doujin Voice illustration included), 9ml, doremi_(doremi4704), hwansang, dydldidl, gulim, hennnachoco, kokosando, misom150, mx2j, myeolchi, nemoga, ronopu, seicoh, yutokamizu, hennnachoco, kitto_ha, ikkia, gotyou, akinashi_yuu, same_(samefgo), jell_(jell_y_fish), etc (unidentified multiple artists who participated in making official Blue Archive illustrations in cutscenes).
Manually removed any drawings that don't match the latest Blue Archive illustration vibes, copyright logo, copyright name, copyright notice, English text, and speech bubble.
Received huge help from the archive, which stores 4K and 8K upscaled images of Blue Archive (big thanks to those who keep the archive running).
Description
FAQ
Comments (11)
I feel like this need a trigger word, the LORA doesn't affect much in my images
I personally don't like using trigger words so this happens with already well-trained subjects
I love it, but Is possible a noob v-pred version?
not at the moment
@BUSANPERSON thanks for the reply
Best gets style very well. Great job waited for someone to get it right. Keep making these for other models or v2? Prompts like flat color and game cg maybe useful but not needed. love your work.
Awesome!
I second a version with a trigger word. The effect isn't as strong even at strength 1.
no trigger words, welp, time to see other models instead
pro tip, the most frequently used tag in the lora is the trigger word and you can check it by viewing metadata. this lora for example, the trigger word would be 'halo' with wooping 605 times tagged in the refined dataset.
Does work of Model WAI-Illustrious-SDXL For V16 Than V14?
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.