This LoRA is at attempt to emulate the Amy Wilkins style of portrait used in some of the Choice of Games interactive fiction games (e.g., Vampire: The Masquerade - Night Road or Werewolf: The Apocalypse - The Book of Hungry Names).Since every image in the training data is of a single character on a simple background, using that composition should lead to the most consistent results. In addition, with such a consistent data set, the more a prompt diverges from a standing completely clothed single character, the more it will need to pull from the base model (which can dilute the effect of the LoRA). <h3 id="tags">Tags</h3><ul><li>amywilkins</li><li>simple background</li></ul>

Files with SHA256: 4dc03a34b9b3fc4b0b15aa5a56f8e214c5899932f17f0f427c1e65452153353d