<h1 id="heading-263">I've created a new and better LORA for wetting. <a target="_blank" rel="ugc" href="https://civarchive.com/models/268341/peeing-clothes-photo">https://civarchive.com/models/268341/peeing-clothes-photo</a></h1>I made it on a new page because it's a new start! This page will no longer be updated. Updates will go on my other wetting LORA.Update for December 7 2023. The ghost of wetting's past, present, and future visited me and told me how to prompt my badly made LORAs to get good wetting images out of them. Here's one example that gives really good wetting images. Note that you'll need to use <a target="_blank" rel="ugc" href="https://civarchive.com/models/132632/epicphotogasm">EpicPhotoGasm</a> or <a target="_blank" rel="ugc" href="https://civarchive.com/models/35960?modelVersionId=204878">Animerge. </a>I don't know why it only works with those.Positive Prompt:<pre><code>still from a movie, cinematic, 20 year old woman, ((pee stain, peeing self)), (pee), looking at viewer &lt;lora:PeeingThemselfV2_0:0.7&gt; &lt;lora:Real_Wetting_Self:0.7&gt; &lt;lora:CombinedPeeingSelf-08:0.5&gt;</code></pre>Negative Prompt:<pre><code>nude, mouth open</code></pre>"Nude" because the model loves to make people naked if you don't specify a type of clothing, "mouth open" because open mouths lead to horror faces sometimes.This style of prompt has gotten me the best results so far. I also found out ControlNet works good with it sometimes too so try that if you want a specific pose.

Peeing pants and other clothes.

<ul><li>Added "pee stain" to the captions in the dataset which now allows the model to make better pee stains. Also fixed some minor incorrect captions.</li><li>Discovered how much of a pain it is to edit text files using Google Drive.</li></ul>It seems like it was only yesterday I released my first pants wetting LORA...oh it was! If you've used it you probably noticed it's not that great at making pee stains. It can do it, but it just doesn't give that massive wet spot us weird fetishists demand. This update now makes large pee stains on most types of clothing. Oddly jeans still exhibit the same behavior as my first version with relatively small pee stains. I can only assume I don't have enough jeans with pee stains in my dataset.To use this version use "pee stain" to get a wetting with a pee stain. Add more attention to "pee stain" to make the pee stain larger. If you don't like the massive gushers this model sometimes produces add "pee" to the negative prompt to reduce the amount of pee while still maintaining the pee stain. If you want a wetting with a small pee stain use "peeing self". If you use "peeing self" adding "pee" to negative prompts will result in no wetting at all.Changing the weight on the LORA will change the size of the pee stain and how much pee there is in general. Normally I use 0.7 weight, but for this LORA 1 weight is working out for me. Don't be afraid to play with the weights though.Just like the previous version I've found that <a target="_blank" rel="ugc" href="https://civitai.com/models/9942/abyssorangemix3-aom3">https://civitai.com/models/9942/abyssorangemix3-aom3</a> works really well with my LORA. Other animated models do work to some degree however.

Files with SHA256: 8798c24addd8626c5f45859dcfd5c25e720a810a47cbef8f12dcd9e903a8f2bb