<p>The purpose of this LoRA is to generate more interesting breast play trained off the Wan2.2 base model. Please be kind in the comments. Constructive criticism is always welcome as it helps others including myself. Also, positive feedback is much appreciated &lt;3</p><hr /><h3 id="version-2"><span style="color:rgb(121, 80, 242)">Version 2</span></h3><hr /><h3 id="features"><span style="color:rgb(250, 176, 5)">Features</span></h3><ul><li><p>Reduced Face changing issues compared to Version 1</p></li><li><p>Additional breast play with another subject involved. Original dataset included, too.</p></li><li><p>Trained on close-up and full body</p></li><li><p>Trained on different style of breast play: grabbing, lifting, covering, pinching nipples, and holding</p></li><li><p>Trained on breast play with single breast play: right breast and left breast</p></li></ul><h3 id="tips"><span style="color:rgb(230, 73, 128)">Tips</span></h3><ul><li><p>Tested on High Str 1 and Low Str 1 to 0.8. Appears to work with 1 strength for both.</p></li><li><p>Increase High Str &gt; 1 to enhance the breast play and overcome issues described below. I tested at 1 and 1.5 for High, works well. Leave Low between 1 and 0.8</p></li><li><p>Prompts</p></li></ul><pre><code>she grabs both her breasts</code></pre><pre><code>she continues to grab both her breasts</code></pre><pre><code>the man is grabbing her breast</code></pre><pre><code>she is covering her breasts</code></pre><pre><code>she rubs her breasts</code></pre><pre><code>she grasps her breasts</code></pre><pre><code>she grabs her breasts with both hands and pinches her nipples with her fingers </code></pre><p>Lucky if you can get this ^^^ one to work, its awesome!</p><ul><li><p>Make sure you are handling the high / low transitions correctly as it relates to the steps. Here is a k sampler that does it well: <a target="_blank" rel="ugc" href="https://github.com/stduhpf/ComfyUI-WanMoeKSampler">https://github.com/stduhpf/ComfyUI-WanMoeKSampler</a></p></li></ul><h3 id="issues"><span style="color:rgb(250, 82, 82)">Issues</span></h3><ul><li><p>Speed up LoRAs may slow down the motion. This is a KNOWN issue with the speed LoRAs. More information: <a target="_blank" rel="ugc" href="https://huggingface.co/lightx2v/Wan2.2-Lightning/discussions/20">https://huggingface.co/lightx2v/Wan2.2-Lightning/discussions/20 </a>and Developer comments here: <a target="_blank" rel="ugc" href="https://huggingface.co/lightx2v/Wan2.2-Lightning/discussions/26">https://huggingface.co/lightx2v/Wan2.2-Lightning/discussions/26</a></p></li><li><p>Version 2's breast play is more subtle then Version 1. I reduced the training Epochs for Version 2 to resolve the face change issue but this seems to have made the breast play more subtle compared to Version 1. Try using V1 high and V2 Low or increasing the V2 High &gt; 1 and keeping V1 at 1.</p></li></ul><p><span style="color:rgb(64, 192, 87)"><strong>Shoutout</strong></span> to all the LoRA creators out there doing the insanely tedious work of tagging, editing and running training sessions</p><hr /><h3 id="version-1"><span style="color:rgb(121, 80, 242)"><strong>Version 1</strong></span></h3><hr /><p><span style="color:rgb(250, 82, 82)">(issue)</span>: Some users experience face changes with this LoRA. This is because the LoRA is overcooked. The LoRA is still usable. However, it means that the strength and workflow may need to be adjusted - review tips.</p><h3 id="features"><span style="color:rgb(250, 176, 5)">Features</span></h3><ul><li><p>Compatible with both T2V &amp; I2V Wan2.2 Models</p></li><li><p>Trained on close-up and full body</p></li><li><p>Trained on different style of breast play: grabbing, lifting, covering, pinching nipples, and holding</p></li><li><p>Trained on breast play with single breast play: right breast and left breast</p></li></ul><h3 id="tips"><span style="color:rgb(230, 73, 128)">Tips</span></h3><ul><li><p>Try between 0.5 to 1 strength. All testing was at 1 strength on I2V with two LoRAs (speed + Breast Play).</p></li><li><p>Reduce strength on the LoRA if the face changes.</p></li><li><p>Consider using <a target="_blank" rel="ugc" href="https://github.com/stduhpf/ComfyUI-WanMoeKSampler">https://github.com/stduhpf/ComfyUI-WanMoeKSampler</a> or adjusting High / Low steps 'correctly' <a target="_blank" rel="ugc" href="https://www.reddit.com/r/StableDiffusion/comments/1mkv9c6/wan22_schedulers_steps_shift_and_noise/?tl=fr">https://www.reddit.com/r/StableDiffusion/comments/1mkv9c6/wan22_schedulers_steps_shift_and_noise/?tl=fr</a></p></li></ul><h3 id="my-wan2.2-cover-videos"><span style="color:rgb(76, 110, 245)">My Wan2.2 Cover Videos</span></h3><p>Workflow is T2I SDXL/Flux/Pony -&gt; Image Upscaler -&gt; Wan2.2 I2V -&gt; Upscaler -&gt; Add FPS</p><p><a target="_blank" rel="ugc" href="https://civarchive.com/models/1916073">https://civarchive.com/models/1916073</a> (text2img &amp; img2vid). Credit DonutsDelivery &amp; Daxamur</p><p><a target="_blank" rel="ugc" href="https://civarchive.com/articles/14582/comfyui-workflow-sdxl-facedetailer-upscaler-and-batch-selection">https://civarchive.com/articles/14582/comfyui-workflow-sdxl-facedetailer-upscaler-and-batch-selection</a> (text2img)</p><h3 id="training"><span style="color:rgb(64, 192, 87)">Training</span></h3><p>Trained on Wan2.2 T2V base model. 100 high quality images and 50 videos that were handpicked and farmed raised to perfection using musubi-tuner. This LoRA works with both I2V and T2V.</p><p>Adjusted training variables so that it works better with other LoRAs. Training was fined-tuned over 10+ runs for each model to get the optimal loss accuracy. I used different tagging descriptions for the dataset to improve variability.</p>

Wan2.2 - Breast Play (I2V)

<p>Less overcooked model then version 1. Should have fixed the face change issues.</p><p>Added more video &amp; images with another person doing the breast grabbing. The core strength of the model is still individual grabbing their own breasts.</p>

Files with SHA256: a3f553bfd996c0c2858e9ce10fd668eadec8d7e9b874bbcd265649d8343e8b1b