<h1 id="ponyxl-version-ml893ipvs"><a target="_blank" rel="ugc" href="https://civarchive.com/models/397525/jin-avatar-the-last-airbender-ponyxl">PonyXL Version</a></h1><p></p><h1 id="base-prompt-dmr1s7ovv">Base prompt</h1><pre><code>1girl, jin, brown hair, twin braids, green eyes, green kimono</code></pre><h2 id="second-hairstyle-frni3beqn">Second hairstyle</h2><pre><code>hair bun</code></pre><p></p><h1 id="tips-and-problems-k9lpdqzjz">Tips &amp; problems</h1><p>The goal of this model is to mimic the original art style as much as possible.</p><p>So the dataset includes only original show screencaps, and since 2/3 of it happens at night and everything is dark, the generated images created may appear faded. Unfortunately, adding <code>night</code> and <code>dark</code> to the negative prompt doesn't really help. Good at generating night scenes</p><p>We can fix it by adding more synthetic images that happens at least indoors and not at night.</p>

Jin Illustrious [Avatar: The Last Airbender]

<ul><li><p>dims 32 alpha 16</p></li><li><p>dataset consists fully original show screencaps</p></li><li><p>98 images</p></li></ul>

Jin_Illustrious-000017.safetensors

1203132_training_data.zip

v2 O

v2 X

<p>Base prompt</p><p>jin, brown hair, twin braids, green eyes, green kimono</p><p></p><p>Second hairstyle</p><p>hair bun</p><p></p>

Files with SHA256: 36b33aaee23bb35ea35a55b03d27a41cce68ab4cc7650cab51d6a99908284fa2