Pony version of the LoRA: <a target="_blank" rel="ugc" href="https://civarchive.com/models/1293938/nova-sentai-ryuseiger-or-giga-tokusatsu-heroine">https://civarchive.com/models/1293938/nova-sentai-ryuseiger-or-giga-tokusatsu-heroine</a>Notes:<ol><li>Training data only consists of aqua and pink. As such, for other colors, you may need to add aqua and/or pink into negative prompts, especially if you are aiming for a color near either or both of them (such as blue or red)</li><li>Due to above reason, the spandex texture may not be as good with other colors, and you may need to increase the weight of the LoRA for the spandex texture. In my own testing, I have increased up to 1.4, though due to the LoRA being trained exclusively on photos, this meant the art style became less anime and more realistic the higher the LoRA weight, though this is a sacrifice I was willing to make since my priority is the spandex texture. In addition, I tried increasing the weight of "shiny clothes, spandex" prompts (both of which were used in the training data of this LoRA), to not much effectiveness</li></ol>Prompts:<pre><code>sentairyuuseiger, white neck, [insert color here!] bodysuit, [insert color here!] skirt, gold belt, [insert color here!] arm guards, white gloves</code></pre>Up to you if you want (personally I do!):<pre><code>shiny clothes, spandex</code></pre>For full body images:<pre><code>white boots, colored shoe soles</code></pre>Helmet:<pre><code>helmet, [insert color here!] helmet, worn helmet</code></pre><pre><code>helmet, [insert color here!] helmet, no headwear, unworn helmet, holding helmet</code></pre><pre><code>white open face balaclava</code></pre><pre><code>no headwear</code></pre>

Files with SHA256: 3eb7ce81348f377272acd20eba087d6e2fd9996a1ce8808b343dbaee5e582fd3