Edit-im still ironing out the kinks. (morph:1.2) seems to work out better. Made this on a whim. its still pretty janky since its really hard to use with negatives because the gradual morphs technically count as different people in the eyes of the AI. in any case.Its a little different. I tried to train it with different tags but it always moves towards a certain way because of the style.Trigger words=animorph, y morphing into x, step-by-step transformation,you really need to lead it though. its should be something like this.<ul><li>animorph, girl, morph, dog, left to right transformation, multiple people, 5 stage transformation, gradual transformation, book cover</li></ul>it should have the subject and the end result. dont try to describe it. if you do need lets say a girl with black hair. it would be something like this ((girl(black hair, glasses)). The pony model is way different than flux so its the best method ive found. lets say you want a female to turn into a female character. the problem would arise that the AI would misunderstand and turn the final character into the final character without the starting process. ill give an example. a female turning into buttercup the pony. because the pony is female it would not understand the instruction. the example i gave you up top should normally be used. just change the starting and ending result.be very careful with using negatives. it really confuses the AI. well because technically its multiple characters with known "stages" that it can consider as deformities. also some negatives you should put in would be book , and exclude people not in stages. you want to include book in negatives because it mistakes book cover as book. so it helps to remove any books and not the book cover style. excluding people not in stages is pretty important since it wants to make the people different for some reason. its not perfect but its okay for the moment. ive also seen that i get results all across the board with different checkpoints. so its hit and miss.and please feed my buzz addiction so i can try and train more loras. thank you. i appreciate any feedback.

Files with SHA256: 97f8061ba68c3b4d337a4e2205fe182037c8285ebee70f66580c2a18649aa5ed