a lot of work was done to create this is lora, trained on all the works of aromasensei ever drawn (except for futa), when generating images it is better to use the deliberate_v2 checkpoint.
how lora works on other checkpoints:
works lora on the deliberate_v2 checkpoint with different Sampling methods:
evolution version lora:
comparison of generated images with the original:
more works can be seen in telegram
If you liked this lorf, please rate and write a review
Description
FAQ
Comments (8)
Amazing LORA!! Thank you.
Thanks for the good words
B-But. Why not the futa? :(
I'm sorry but I'm not interested
>trained on all the works of aromasensei ever drawn (except for futa)
How many works is that?
a lot of work was done in the dataset, I used about 1500 pictures that were manually processed by me in Photoshop and descriptions were given, but I can’t say exactly how many of them are unique, because in order to get a good result, one picture could have several options (with a background , no background, face only, etc.)
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.