<h1 id="this-is-a-stomach-bulge-lora-for-moss-image">This is a stomach bulge lora for MOSS Image</h1><h2 id="trained-on-300-mainly-comic-and-cartoon-images">Trained on 300 mainly comic and cartoon images</h2><h2 id="edit-for-v2">Edit for V2</h2>I did recaption all the images since i did find a better way of doing it and had the idea of giving it some modular setup. This kind of worked. Those weird 90° dick snap angles have been reduced too.<h3 id="v2-prompting">V2 Prompting</h3>Your main trigger "phrase" is still stomach bulge without it nothing much will happen, unless you maybe describe the effect in different words, but it will not work as well. I did then set up a raster if you want for the captioning:Size: - If Small: "A slight stomach bulge. "- If Medium: "A noticeable stomach bulge. "- If Large: "A deep smooth bulge. "- If Very Large/Long: "An extreme elongated stomach bulge. "and Form: smoothdistinguishedjaggedyou can combine the two freely! smallest /tamest would be A slight smooth stomach bulgebiggest/nastiest would be An extreme elongated jagged stomach bulge<h3 id="you-will-have-to-experiment-a-bit-to-find-what-you-want.">You will have to experiment a bit to find what you want. </h3>As usual prompt and especially other loras and checkpoints do have a strong effect on the output. Have fun.<h2 id="end-edit-for-v2">End edit for V2</h2>It works with just base MOSS but if you want some better prompt adherence or to add fancy things i guess you better use some finetunes or what every else you want to add. it does work with other trained checkpoints of coarse, but you might have to adjust the stomach bulge lora strength. depending on results. it seems to work well with my AIO pose lora.it has picked up a lot of sperm and vaginal fluid in the training process.With negative prompts those things can be reduced.I am not perfectly happy with the bulge it seems to be to round for my taste but it works.... i might try to recaption the whole set and try again.<h2 id="pleas-post-images-you-generated-with-the-lora-in-image-section.">Pleas post images you generated with the lora in image section.</h2>For one of the first loras i ever trained it seems to be working fine. at least for my taste.Pleas like it and do all the usual things ppl request from their loras..... what ever that may be.The uploaded images should contain the used metadata from comfyui.I do not use any other image generating software so on other platforms i will probably not know.Please ask questions if you have them.

EK's Stomach Bulge for MOSS

Recaptioned the images. set them up in stages and shapes. works, not perfect

Files with SHA256: ee22b47e9fbe104eef384832e824ce78360aafabd87b17bfe91a98e7b193c839