A LoRA for creating a pose where a woman holds an oblong object like it's a dick. Training data is mostly 512x768. I've found that lowering the fruit/vegetable weight to 0.8-0.9 helps to prevent tons of them from being added to the image. SD 1.5 models have trouble with hands, as you know, so it may take a few tries to get a good looking image. Weights between 0.8-0.9 seem to work best, though I've found that lowering the LoRA weight by 0.05-0.1 when using Realistic Vision models leads to better results.Reliable foods:<ul><li>banana</li><li>eggplant</li><li>cucumber</li><li>carrot*</li></ul>*For some reason, "carrot" produces weird results unless you lower the weight to around (carrot:0.8). Even then carrots are weird.Example prompt: a photo of a naked woman holding (a banana:0.9), immodest_muff &lt;lora:immodest_muff:0.8&gt;

Files with SHA256: d57793a9b6d7fc207a2397a02012bac2fc0560bdbde59aadff3552d914e73eda