<h1 id="harness-gag-jrpn44ylv">Harness gag</h1><h3 id="hunyuan-model-v2mywmjxd">Hunyuan model</h3>Trained a Hunyuan video model using the same dataset used for the Flux model (only images).Trained form 1000 epochs using default parameters of diffusion-pipe.A typical prompt should include something like:[something] wearing a red harness gagSometimes only the gag ball will be seen in the mouth and now straps over the head. In that case, you can either increase the strength a bit (to 1.2), or you can add:[something] wearing a red harness gag with black straps over the head<h3 id="flux-model-d6to0o78q">Flux model</h3>Trained on 48 images for 2500 steps.How to use[something] wearing a red harness gag, harnessgagTips<ul><li>Use "close-up photo of" for a zoomed in picture.</li><li>Choose between "from the front, from the side, from above, from below, from the front at an angle" for the direction.</li></ul>

Files with SHA256: dfe0b670c9ee56ac1b9061eda30746aef1c86e2d4984dceca360d9fcf7d83970