<h2 id="krea-2-version">Krea-2 version</h2><p>80% same dataset.</p><p>requires more prompting in human language, simple tags work poorly.</p><h2 id="zit-version">ZIT version</h2><p>Trained on 247 images.</p><p>For women in panties, use lower values ​​(0.5 - 0.7) since most of the dataset is without panties.</p><p>For a lower viewing angle, it is better to use a landscape aspect ratio, and for a high angle, a portrait one. (tag words: low angle, mid angle, hi angle. But this doesn't work very well.)</p><p>Trained on comma separated tags.</p><p>Works on base checkpoint.</p><p>If a girl is lying upside down, try adding "in the foreground is the huge ass of a lying girl."</p>

Files with SHA256: e0f7e10d565f71228499181c2cd103823efdd1db1a3516381f8b622b349773c6