Smeargle LoRA trained on Pony upon request. Works incredibly well with a few issues, which I'll note below: Issues:<ul><li>Smeargle's tail can be missing the paint color on the tip at times<ul><li>Add 'green tail tip' or whatever color you want to your prompt to improve the tail consistency</li></ul></li><li>Smeargle can come out with multiple tails<ul><li>Add 'multiple tails' to your negative to improve tail consistency</li></ul></li><li>Smeargle's fur collar can sometimes be interpreted by your base model as a regular collar, and have a tag put on it<ul><li>Add 'collar' to your negative if you run into this issue</li></ul></li></ul>

Files with SHA256: c781a55bf27694558fe5871cd12c934e287360588557222fbddea116fe9358af