A general "Video Game" / 3D model style foundation LoRA. Able to get acceptable results with multiple Pony models. Pony Models I Found Success With So Far: <a target="_blank" rel="ugc" href="https://civarchive.com/models/257749/pony-diffusion-v6-xl?modelVersionId=290640">Pony Diffusion V6 XL</a> + <a target="_blank" rel="ugc" href="https://civarchive.com/models/359876/blendermix?modelVersionId=402216">BlenderMix</a> + <a target="_blank" rel="ugc" href="https://civarchive.com/models/341433/everclear-pny-by-zovya?modelVersionId=511677">Everclear PNY by Zovya</a>. Bear in mind though, that it did seem that given the extra dimension, that prompts given to the current model one is using for image generation, whether on the positive or negative side would have the references alter the results using a "realistic" source image as <a target="_blank" rel="ugc" href="http://Civit.AI">Civit.AI</a> tags it or alternatively to the side of "Anime" / "2D" / "Toy". After understanding if the image generated was giving me 2.5D - 3D results before I applied this LoRA, allowed me to weigh the additional styles or different Pony model variants to alter the image with more consistent results. Additional Support LoRA : Styles For Pony Diffusion V6 X - 2.5D Realistic (Both Version <a target="_blank" rel="ugc" href="https://civarchive.com/models/297619?modelVersionId=349828">One </a>&amp; <a target="_blank" rel="ugc" href="https://civarchive.com/models/297619?modelVersionId=422554">Two</a>)Old Time Post Apocalyptic Enviroment XL - <a target="_blank" rel="ugc" href="https://civarchive.com/models/120651?modelVersionId=151897">v3.0-Chernobyl</a>Negative Embedding for Realistic Vision v2.0 - <a target="_blank" rel="ugc" href="https://civarchive.com/models/36070?modelVersionId=42247">v1.0</a> These tools helped without having to outright swap which model I was using to impact the degree of "3D depth" I could get, and also not having to lose too much progress. -Although the foundation is completed. The list of recommended / successful Pony variant models and additional support LoRA will continue to be updated as I try out the other LoRAs / Pony models.

3D Style : Pony LoRA

Default : 1.0Range : 0.0 - 4.0<ul><li>Generally I found 3.0-3.5 as the limit with the focus of the image being a person, but could go to 4.0 for some environments / landscapes that still performed correctly.</li></ul>

Files with SHA256: bc01d342bcdb18cf6e431374a9f9dddc3ff561ac01b1bfa91d3036d58e850305