<p>Trained on 38 fanart images. similarly to the John model, it was surprisingly hard to find good training images. the godtier and non godtier prompt are meant to be used separately- the non-godtier prompt doesn't seem to work as well as the godtier prompt, but using them both at the same time seems to be able to give a better non-godtier version at the cost of accuracy. Turning the weight down also helps the non-godtier version at the cost of accuracy. Trigger prompts should be self explanatory otherwise.<br />It's also somewhat hard to change the style with a lot of loras and models</p>

Files with SHA256: 97758a9edc54c16ba8794ae9b07c747a5459f5749838ba3f35960f764e1ccd41