Second version is not necessarily just an upgrade, they are slightly different
Trained using retro-pop art-lofi style pictures, with the goal of creating colorful images with a kawaii aesthetic
The perfect weight may vary due to checkpoint, recommended weight somewhere around 0.7 and 1.0
Description
FAQ
Comments (12)
whose brain did read "retro loli" at first, too?
They call it "mind in the gutter", I think :D
It would have happened to me if I hadn't been the one who put the title
@victorc25744 more like they call it illegal, since the term refers to illegal inhumane thoughts about children. tsk tsk tsk.
@SOLTIS get some help
this was meant as a fun comment... some people really just should bleach their brains to get them fresh and clean again...
Yea I have no idea how to use this.. setting the lora strength to 1 and using the trigger words, it doesn't get anything close to preview images.
Why not upload a prompt to show an example?
@OrangePumpking the main issue is having no prompts and the specific model to use to reproduce results. Under those circumstances, most LoRAs are sadly useless
@OrangePumpking If you strip meta data from your LORA, that's okay. It's up to you - I'm not going to complain.
My issue is your preview images also have no meta data. We have no idea how to use the things you spent so much time making.
@DreamExplorer Metadata added for the preview pics, checkpoint is Yotta Mix
@OrangePumpking You're the best ~ thank you so much for your work & efforts.. will check it out!
@DreamExplorer No problem, If you guys have another question let me know