<h1 id="heading-58">LoRA for the Style used by Artist Vivivoovoo.</h1><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/51256d24-4fcf-44b1-81c6-ffc5d2298bf7/width=525/51256d24-4fcf-44b1-81c6-ffc5d2298bf7.jpeg" /><a target="_blank" rel="ugc" href="https://ko-fi.com/reeveeai/commissions">If you like my work and want to support me, please consider supporting me on my Ko-fi page. I'm open to donations and commissions for art and LoRAs with no censorship. This text is a link by the way :)</a>I trained it this with approximately 220 pics. I trained it at 15 epochs with a dim/alpha of 32/16.Not a requirement, but I would appreciate referencing me if you use my LoRAThis LoRA helps style females in the art style of Vivivoovoo. I am uncertain how popular they are in general but I very much like the colour composition used in all their artworks and facial expressions and highly suggest searching up their work. The effect is somewhat subtle in certain cases, but more distinct at higher LoRA weights. Let me know how it goes for you. Read stuff below on guidelines and known issues :)<h1 id="heading-60">Triggers:</h1><ul><li>Input "Vivivoovoo" near the start of the prompt.</li><li>Input different colour hair, eyes, skin tone, hair style etc and you should be able to create original characters quite easily.</li><li>It is good at handling prompts that involve women smiling, looking down at viewer etc. Refer to reference pictures for examples.</li></ul><h1 id="heading-60">Prompting Tips:</h1><ul><li>To get this to style a character LoRA you may have to reduce this LoRA weight to something below 1, I used 0.9 and the results seemed sufficient when used with character LoRAs that had their weight set to around 0.6 or 0.7.</li><li>Due to the training data, this works quite well with other Pokémon character LoRAs namely the older women such as Cynthia, Lusamine, Sabrina, Elesa etc.</li></ul><h1 id="heading-61">Notes:</h1><ul><li>My experience has shown scale 0.7 ~ 1.2 works best; my best results were 1.</li><li>If combining this with other character LoRAs consider reducing the weight of this LoRA to 0.9 or less.</li><li>My prompt order shown in the sample pictures worked best for me.</li><li>Works on different models though weights may need to change for good results.</li><li>I tested this on the AnyLora model, Foxyansfw and Midnight Maple model. Midnight Maple gave best results for this LoRA for me.</li><li>Refer to my pictures for examples and what upscaler I used, keep in mind I have used other textual inversions to help with poses and negatives.</li><li>I think this turned out well at least for the creation of original characters or some Pokémon characters. If there seems to be a high demand for a new version that more flexible then I will consider updating this.</li></ul>I don't think people read entire descriptions like this, but congrats if you do

Files with SHA256: 6e0002ed660f1dc1f9a8dab6bbab9872aade6b2cca1d694c10477f466bf6920b