<h2 id="heading-7">Each "version" above is a different character, each character is a separate LoRA.</h2> <a target="_blank" rel="ugc" href="https://ko-fi.com/h4rv3yd3nt">Beer and coffee keep my motivation up</a> 🍻The activation tag for each is "[character name] + sdv" aka "alex sdv", "harvey sdv", etc. Some notes about these LoRA's ↓- Each LoRA was trained with about 40 images. Because of the original pixelated graphics, there's some debate about specific characteristics such as eye or hair color in cases where it wasn't clear. For example, whether or not Shane and Sebastian have purple hair or if their hair is black and the purple we perceive is a stylistic choice to indicate lighting. Those details are in your hands, specify the color you want to see, they both were tagged and included in the training data. - In the case of Harvey, I did not tag his eyeglasses or his mustache because during testing I found the character name alone didn't result in a good likeness. You could probably remove them with those tags or similar in the negative prompt. - In the case of Shane, I think he looks best when you add 'stubble' or 'facial hair' to the prompt. - And finally, just like every other safetensor LoRA out there, you can open them in a text document and see the complete list of tags they were trained on in the metadata header. Be warned, a bit of lewd data was included in each, Alex had by far the most. This will help you get their typical outfit if you want it though.If you use this LoRA as part of a mix or host it on a generation service, please mention and link back to this page (especially if you're making money off of it)

Bachelor's of Stardew Valley (6 LoRA Pack)

<ul><li>~40 images</li><li>Booru tagging</li><li>x512 base resolution with aspect ratio bucketing</li><li>100 epochs, batch size 2</li><li>DAdaptAdam optimizer</li><li>32 net dim, 16 alpha dim</li><li>Clip level 2</li><li>[redacted] base model</li></ul>

Files with SHA256: c230b14a65c8d421a2d167224974857ead2636f563cb708b3746ebe95d59f231