<p>Everyones favorite tanned dragon girl that may or may not be actually a boy.</p><p>If you like the LoRA, please leave a reaction, review and/or tip. Enjoy!</p><p></p><h3 id="heading-7867">Usage</h3><ul><li><p>Trigger word: <code>adcvr</code></p></li><li><p>Support tags (only use if required): <br /><code>tan, dragon horns, dragon tail, medium hair, black hair, red hair, multicolored hair</code></p></li><li><p>Recommended strength 0.7 - 0.9 (for lite: 0.8 - 1.0)</p></li><li><p>Use Clip skip=1 or Clip skip=2 (other values may also work, use at own risk)</p></li><li><p>Works best at 1024x1024. Similar resolutions (e.g. 800x1200) also work well.</p></li><li><p>Some tags may have strong style bias, because of small training dataset (for example: <code>gris swimsuit</code>). This can be overcome by explicitly specifying an artist style, either via LoRA or <code>by &lt;artist&gt;</code> tag.</p></li></ul><p></p><h3 id="heading-7868">Example</h3><p>See preview images (generated with pruned version)</p><p></p><h3 id="heading-7869">Unpruned vs. Pruned vs. Lite</h3><p>I offer three variations of the model, which only differ in the file size and quality. Pick the one that best suits your preferences. If you can not decide, simply pick the "pruned" version. Note that the quality is <em>relative</em>, even the lite version usually produces images which closely resemble the intended subject.</p><ul><li><p>unpruned: The originally trained model. Largest file size, best quality. Pick if you want the best possible quality, at the cost of disk space.</p></li><li><p>pruned: Downsized model. Almost same quality as the unpruned version, with noticably smaller file size.</p></li><li><p>lite: Drastically downsized model. Smallest file size. Still has good quality, but noticeably worse than unpruned and pruned.</p></li></ul><p></p><h3 id="heading-7870">Notes</h3><p>My first character LoRA. The dataset for this one was kind of small, but I am happy with the result. I might experiment with extending the training data using "artificial" data (i.e. generated images instead of hand drawn ones), but I think don't think the quality will improve much.<br />More character LoRAs are comping up.</p><p></p><h3 id="heading-510">Using training data</h3><p>Feel free to contact me if you want access to the training data.<br />The training data (images and tags) have been mostly sourced from <a target="_blank" rel="ugc" href="http://rule34.xxx">rule34.xxx</a> and other images boards. They have been cleaned up manually by me. When using the training data or modified versions of it to create models or datasets, please credit me as appropriate.</p><p></p><h3 id="heading-7871">Disclaimer</h3><p>This LoRA is not affiliated with the character it is based on. If you like the character, check them out at one of their socials:</p><ul><li><p><a target="_blank" rel="ugc" href="https://www.youtube.com/c/adcvr">https://www.youtube.com/c/adcvr</a></p></li><li><p><a target="_blank" rel="ugc" href="https://www.twitch.tv/adc_vr">https://www.twitch.tv/adc_vr</a></p></li><li><p><a target="_blank" rel="ugc" href="https://twitter.com/adc_vr">https://twitter.com/adc_vr</a></p></li></ul>

ADC Vr (character) [PonyXL]

<p>This is the pruned version. It is significantly smaller than the unpruned version while loosing a negligible amount of quality.</p><p>This is the recommended version.</p>

Files with SHA256: c7ee0528d900ac66a8a1cd3df6aad8e8904b5eeb18b523cc5541865f4b250a57