<h2 id="loras-junkyard">LoRAs Junkyard🗑</h2>so what this is exactly? as the name suggests its a junkyard post for all the LoRAsthat I have lying around but didn't bother posting .. stuff like:<ul><li>failed experiments </li><li>quick proof of concepts LoRAs</li><li>bad quality LoRAs</li><li>LoRAs that i forgot i had lying around</li></ul>I figured out that "bad" is subjective and someone else might actually enjoy having these around.<ul><li>for use tips , I might include some in the version description </li></ul><ul><li>why a single post ? things can get messy over here (unordered posts, old version ... ) to I don't want to deal with that mess over my whole gallery</li></ul>

LoRAs Junkyard

This was follow up experiment to the <a target="_blank" rel="ugc" href="https://civitai.com/models/2239703?modelVersionId=2544837">compound LoRA experiment</a> , where i kept the same dataset but unified all under the same Trigger Word ' compd-d', the overall result was weaker then expected and failed to produce the intended sub-style nor consistent mixed style like the 1st one.but I be keeping it in ma collection as it got a subtle style to it.

Files with SHA256: 333167b376ad59e00c3f043470e3fe285079c940459a94e04e03f834d5873209