<p>from anime ひなろじ~from Luck &amp; Logic~</p><h2 id="for-nlp-model">for NLP model</h2><p>try to description like this:</p><pre><code>riones, it's a full body shot of a girl. She has long blonde hair and green eyes. Her breasts are small. She seems to be crying and tears on her face. She is wearing magical girl outfit. Her magical girl outfit looks dirty. She wears a skirt decorated with pink frills and pink roses. A white leotard with pink frills and green stripe is worn under the skirt. She wearrs pink high heels decorated with footwear flower and leg ribbon. On her neck,she is wearing a pink flower choker with neck ribbon,and a short necktie also. arm warmers with pink frills. She wears green hairband,hair flower and bridal veil . She also wears earrings.</code></pre><p>I am not testing it with tag words, but I think it's also work.</p><p>You could try yourself. Just describe the parts of her outfits like what I show above.</p><p>PS: actually, model trained by NLP looks more flexable for the prompts</p><p>Welcome to give me any feedback and upload your result.</p><p></p><h2 id="illustrious">illustrious</h2><p>weight about 0.8</p><pre><code>riones,bare shoulders,strapless leotard,white leotard,pink skirt,dress flower,white arm warmers,leg ribbon,ankle boots,high heel boots,pink footwear,green hairband,hair flower,pink choker,short necktie,very long hair,blonde hair,green eyes</code></pre><h2 id="sd1.5">SD1.5</h2><p>weight 0.6~0.8</p><pre><code>rionesu, high heels, pink footwear, hair ornament, hairband, hair flower, thorns</code></pre><p></p>

Liones Yelistratova

<p>The model is trained by NLP captions. It's just a test for training by this way. It seems ok.</p>

Files with SHA256: 7ec43f039e3d29f7f7aa53dac82d3e69b9eb039fa5d7e22ea7374a80ed874e70