These lora were trained with 50-500 images from the game Rance 03 and some of 02. The idea was to capture the style that Onono Imoko used in the game. They can reproduce the GC style from the game's events as well as the full body character illustrations. (it was all in the data set used)
To get the results you see in the images above, these are the settings I used.
V4 was trained on 50 images using specific settings for vpred. The base for this lora is Noob 1.0 V-Pred.
V3 was trained on 50 images using improved training settings. The base for this lora is Zootlustrious V2.
V2 Illustrious 0.1 was trained using 50 selected images from the V2.0. Please use the prompt from the example images for the most optimum results.
V2 Hadrian DeliceXL 2.0 was trained using 50 selected images from the V2.0. I haven't tested the lora on any other pony based model, but I speculate it won't work very well on anything but Hadrian. This version sits nicely in between the AustimMix and Ponymagine lora. It looks much closer to the source material like Ponymagine but has the prompt freedom of the AustimMix. I highly recommend you use this version.
The V2 of Ponymagine was trained using 50 selected images from V2.0. I also tested it on Animagine XL V3.1 and it seems to work (barely) but I highly recommend you just use it on Ponymagine since it's the entire reason I made this version. I do not guarantee support for anything else. Please read this if you're wondering why I made this.
The base for V1 uses Pony Diffusion V6 XL and it works on AutismMix as well. It also produces great images. Choose this one if you work exclusively in base Pony.
Description
FAQ
Comments (11)
Man, keep cooking! Example pics seem very promising, I'll try it later, need sleep - pony is very addictive thing :D
Pony is fantastic. It does hands and feet very well compared to any other stable diffusion model. That is why I train on it. I will continue to do what I can.
This lora is amazing! Have you consider to make a version for AnimagineV3?
I've thought about it, I wonder how it would turn out. It takes me a night to train something like this and I have a few other things I want to train so I'll try it out once that's all done.
Did you use only single tag for all images during training?
"ss_tag_frequency": { "20_CoolRance style": { "coolrance style": 463 } },No, the frequency tags from the text file weren't included in the metadata. I wasn't aware of the option when I trained this lora, I'll try to include them from this point forward.
"The idea was to capture the style that Onono Imoko used in the game."
Is style of arts in game same as style of his arts on danbooru?
A lot of their recent art on danbooru is more colorful and stylized like an illustrator's work. I wanted to capture the style reflective of the Rance game's CG art.
how do you add the high res fix to the workflow? Do I need to download any additional addons for it?