Tools that I used to make this LoRA: <a target="_blank" rel="ugc" href="https://github.com/mifi/lossless-cut">LosslessCut</a> (To capture PNG images from source), XnView MP (For image cropping), <a target="_blank" rel="ugc" href="https://github.com/Jelosus2/DatasetEditor">DatasetEditor</a> (for captioning), <a target="_blank" rel="ugc" href="https://github.com/derrian-distro/LoRA_Easy_Training_Scripts">LoRA_Easy_Training_Scripts</a> (for training).<h1 id="tags:">Tags:</h1><h3 id="style:">Style:</h3>Remove this tag, or decrease its weight if you need better style flexibility<pre><code>anime screencap,</code></pre><h3 id="appearance:">Appearance:</h3><pre><code>ankouguisumv, 1girl, solo, black eyes, brown hair, round eyewear,</code></pre><h3 id="haircuts:">Haircuts:</h3>Short ponytail:<pre><code>medium hair, sidelocks, short ponytail, folded ponytail, </code></pre>Long ponytail:<pre><code>medium hair, sidelocks, ponytail, </code></pre>Hair down:<pre><code>medium hair, hair down, </code></pre>Short hair:<pre><code>short hair, </code></pre>*There is a scene in restaurant where her hair color is of a different brown shade, I tagged them with alternate hair color tag, it doesn't have strong effect in positive prompt, but adding it to negative prompt helps with getting more consistent hair color<h3 id="outfit:">Outfit:</h3><pre><code>brown coat, open coat, </code></pre><pre><code>brown jacket, open jacket, </code></pre>+<pre><code>grey shirt, turtleneck shirt, shirt tucked in, black skirt, partially fingerless gloves, brown footwear, high heel boots,</code></pre>

Files with SHA256: 53f6de3a50c5b2ab156da6eb94a6bd68635a7630ecc3213c3be23fbf3540b451