<h1 id="vampire-contessa-(oc)">Vampire Contessa (OC)</h1><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/922f7b79-d56a-43e9-bed7-57e25197b0fd/original=true/922f7b79-d56a-43e9-bed7-57e25197b0fd.jpeg" />This one is a new OC I started working on after accidentally stumbling across her while testing other models. I'm sure it pulled parts from characters in the illustrious training data, but I thought she was interesting enough to keep. I trained with images in different styles for traditional media and digital art, to give a good blend and prevent overtraining a specific style.There are three main tags for this model, "Contessa" for the character herself, "Vdress" for her dress, and "Vsleeves" for detached sleeves.The "Contessa" tag should give the character with pale skin, long blonde hair, red eyes, fangs, pointy ears, and the head wings, automatically without having to prompt those traits.<h3 id="vdress-and-vsleeves">VDress &amp; VSleeves</h3><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/b9efacb4-2384-4b8c-bd64-837d96c1f235/original=true/b9efacb4-2384-4b8c-bd64-837d96c1f235.jpeg" />Sometimes just prompting either "Vdress" or "Vsleeves" will give you both the dress and the detached sleeves, but usually leaving either out will remove them. If you get the sleeves without prompting, using "sleeveless" in the positive prompt should remove them, likewise using another clothing type, or the "topless" or "nude" tags should remove the dress. As for color, usually using these tags with the "Contessa" character tag gives a black or blue dress with blue highlights, though you may get random colors like white or red without a specific dress color tag. I've had luck with putting the color before the "Vdress" or "Vsleeve" tag, as well as using a separate color tag; e.g. "pink vdress", "black dress", "red sleeves". (Usually the sleeves default to whatever color is used for the dress).<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/d2c8ea24-229a-49b5-a99c-1b5d699519ff/original=true/d2c8ea24-229a-49b5-a99c-1b5d699519ff.jpeg" />You can also try the "&lt;color&gt; inner dress" tag that was trained (replace &lt;color&gt; with your color choice). This gives a separate color to the inner lining of the dress, though it was only trained with red it should work with any color. You will likely need to include the base dress color as well, and there will likely be some color cross-over.<h3 id="ornate-pasties">Ornate Pasties</h3><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/f65c7678-4dec-4eb5-93be-4751d34c31a1/original=true/f65c7678-4dec-4eb5-93be-4751d34c31a1.jpeg" />I also added two types of special pasties, using three other tags, "ornate pasties", "black bat pasties" and "gold cross pasties". The ornate tag is used with both of the two types to help give them more of a jeweled look. Though these don't have as much training data, so they probably won't be as consistent. <h3 id="the-costume">The Costume</h3><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/1abad616-d230-434a-9311-9ec8d32cccbd/original=true/1abad616-d230-434a-9311-9ec8d32cccbd.jpeg" />I did also include a few images of other female characters with different skin tones and body types wearing the Contessa outfit, and from my testing it seems to work fine. Though you will likely find more inconsistent outputs, and you may need to increase the tag strength for "VDress" and "VSleeves" (for the sleeves you might need to add "detached sleeves" to the positive as well, as I found that it connects them more often when used for other characters). You may also need to add "navel" to the positive to get the chest cutout area to extend down further.The body types will still lean more toward the curvy/thick body of the main character, Contessa, though.Finale Notes:While most of my testing seems to have output nice images, this lora is not perfect. You will need to play with the weights of tags, and find the right combinations that work for you. <span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">These Loras are trained using Illustrious v2, and the images are generated using Fooocus with Wai Illustrious. I did upscale the sample images, and inpaint details (hands and faces). <span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">As with all of my models, you may use them in your own training, merges, or anything else.

Files with SHA256: 2c4454136e4be02d8518ffc580c7558fbdfdb8ac34a308154af8052b7b8609eb