<h1 id="synthetic-female-character-uslms*"><span style="color:rgb(250, 176, 5)">Synthetic Female Character - USLMS*</span></h1><h2 id="model-information:"><span style="color:rgb(34, 139, 230)">Model Information:</span></h2><p>This is a character model I trained just to see if I could create a consistent character from images generated using text-to-image with my RAP checkpoint. As such, this LoRA in not based on any images of an actual human. It is intended to produce a consistent anime to semi realistic likeness of a 20+ year-old female with long brown hair and brown eyes.</p><p>If you find this model useful please post images.</p><h2 id="lora-settings-(sd1.5-illustrious-krea2):"><span style="color:rgb(34, 139, 230)">LoRA Settings (SD1.5, Illustrious, Krea2):</span></h2><h3 id="animesemi-realistic-checkpoints"><span style="color:rgb(230, 73, 128)">Anime/Semi Realistic Checkpoints</span></h3><p>CLIP SKIP: 2</p><p>TOKENS: uslms, 20yo, 1girl, brown hair, brown eyes, makeup</p><p>WEIGHT: 0.75-0.85</p><ul><li><p>Here are some recommended weights based on SD1.5 checkpoints:</p><ul><li><p>Honey2d = 0.75</p></li><li><p>SexyToon3D, Dreamshaper8 = 0.8</p></li><li><p>RealCartoon25D, RealCartoonRealistic17 =0.85</p></li></ul></li></ul><h3 id="(alternative)-realistic-checkpoints:"><span style="color:rgb(230, 73, 128)">(Alternative) Realistic Checkpoints:</span></h3><p>CLIP SKIP: 1 or 2</p><p>TOKENS: uslms, a 20yo woman with long brown hair and brown eyes wearing makeup</p><p>WEIGHT: 0.75-0.85</p><p></p><p><span style="color:rgb(250, 176, 5)"><strong>*</strong></span><strong>Note: </strong>The name USLMS comes from a string of tokens I used to generate the original images used to train the model: american, lisa, michelle, sarah.</p>

Synthetic Female Character - USLMS

<p>Trained on same anime dataset with Krea2</p><p>Flexible style and works with other LoRAs</p>

Files with SHA256: be34baf02d2e2ae8131b8d4c4385679b44fc055f35da0a883e53024be0a6610e