Hi there, this time i am sharing with you my Alice Model from "Solo Max-Level Newbe".This model objective is to help when trying to generate images of alice and i think it does a decent job in doing so.It's not heavely trained so it hasn't a strong style imprinting and it will keep the model style (as you can see by the tests).The activation keyword is:<pre><code>thg-alice</code></pre>This lora work best at around strenght: at a strenght between 0.8 and 1.I almost always use strenght 1.There are other tags i suggest you to use:<pre><code>long hair &lt;-- optional, but some times fix images where the seed try to produce short hairs when nothing is specified.

white hair &lt;-- suggested to use always, cause it limit the color bleeding of other tags in the prompt on hairs

red eyes &lt;-- this tag is optional but as for the hair it might fix some color bleeding from other tags</code></pre>This lora has no particoular biases, except the fact that when you use the tag "red eyes" you might need to place a "BREAK" between character tags and clothes tags to ensure there's no bleeding.It was tested on few different models and it work as intended.The images in the gallery were produce using just highresfix and being upscaled with "4x-animesharp" with denoising strenght 0.2, 10 steps and the multiplier to x2.(So if you really want you can produce much better results, by using things like controlnet or adetailer)As always if you like this lora you can support me by sharing your work, leaving a review, reacting to images in this post or by tipping buzzs.

Alice Von Ataraxia - Solo Max-Level Newbe

This Version is aimed to help when producing images of alice in her "vampire" form.Disclaimer: this version is somewhat unstable, so you might need few tweaks each time to get the image right.It's suggested to use strenght 1 when working with this version (lower strenghts could mess up your hands).Generally you don't need much prompt to generate a good image, you can simply use the activation tag:<pre><code>thg-alice</code></pre>and then follow up with clothes, pose and camera angle.The model doesn't require the "pointy ear" tag, or the "vampire tag", but sometimes you might need to use those to fix some issue in the generation.Other tags that are not mandatory, but that could help are: "long hair, white hair, red hair".For the classic outfit you can generate something similar using "black_dress, cleavage cutout".Generally speaking i'd suggest you to use the "human" version and then prompt from there, but if you want a plug and play lora, welp this might be the one of your choise.Note: It isn't intended to support "bat wings"

Files with SHA256: e01920a7f596dceb2351aec23d7bec1815d624f1ad0623692abe4585cbd5fa3e