<h1 id="what-is-this-ipmt5eciu"><strong><u>What is this ?</u></strong></h1><p>Well basically instead of merging lora (method tested <a target="_blank" rel="ugc" href="https://civarchive.com/models/707873/kinky-merges-style-pony-v6-xl">here </a>on my other lora page) to get an unique style. This time i tried to merge and mix directly from the dataset. Then trained it. And this is the result.</p><p></p><h1 id="how-to-use-z6ay3po7w"><strong><u>How to use ?</u></strong></h1><p>Well it's a lora style. So no trigger word. Just prompt for it and mess with the weigths. I recommend roughly between 0.5 and 0.8. Below it doesn't show enough effect and above it deep fry real quick.</p>

Files with SHA256: fa531e7e91d5555799d0a6185414a6ee861df55af23406c5cdb0adaf5ffd3464