If you enjoyed this model, why don't you check out my other works? I put a lot of effort into correctly recreating artist styles and making the occasional character LoRA.😊This model was trained using the collab/guide by holostrawberry <a target="_blank" rel="ugc" href="https://civarchive.com/models/22530/guide-make-your-own-loras-easy-and-free">[Guide] Make your own Loras, easy and free - colabs | Stable Diffusion Other | Civitai</a>The generation was selected based on testing in AOM3. I recommend experimenting with the weights when using other checkpoints.<h2 id="heading-226">About the artist:</h2>Hanabushi makes highly stylized art and is most well known for making the art of the following music video:<div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/Atvsg_zogxo" start="0"></iframe></div>Check out the artist here:Twitter: <a target="_blank" rel="ugc" href="https://twitter.com/hanabushi_?lang=en">はなぶし🚀KUNG-FU-PIGGY (@hanabushi_) / Twitter</a>Youtube: <a target="_blank" rel="ugc" href="https://www.youtube.com/c/KUNGFUPIGGY_">KUNG-FU-PIGGY - YouTube</a>Instagram: <a target="_blank" rel="ugc" href="https://www.instagram.com/hanabushi_/">はなぶし✨KUNG-FU-PIGGY💫🚀 (@hanabushi_) on Instagram</a><h2 id="heading-227">Weights &amp; Prompts</h2><h3 id="heading-884">AOM3 volatile:</h3>I recommend using weights of 0.8-1.0, adjusted to taste. Check the XYZ plots for a detailed breakdown. This version was specifically trained on AOM3 and probably won't work nearly as well on any other checkpoint.It can really struggle with anatomy, hands and teeth (especially when depicting full bodies) so I recommend using more than just a few seeds per prompt. As you can see in the examples the range between the good and the "bad" is pretty big.Unlike v1.0 this version is a lot better at depicting prompted backgrounds and landscapes in the right style. Setting the weight to .9 tends to be enough to reduce the random semi japanese symbols.<h3 id="heading-884">v1.0:</h3>For me it worked with a weight of 1.0. As usual I have included XYZ plots for reference.Some useful prompts:color contrast, limited palette, flat colorAs a small sidenote, it looks like stable diffusion picks up on the "piggy" in the prompt and will occassionally put pigs into landscapes.Please leave a review if you enjoyed this model or have problems with it. This helps me with improving future models and finetuning them in general. Give it a try and throw some images down below! 👇

Files with SHA256: 25f398255ada658d23c74c4a5ea50db99cdd59405919536f965cbbc2a6a263c5