Tried training a LoRA on a Funko Pop, just to see how flexible it would be. Between the limited data set, and the singular pose, it isn't all that flexible with default weights. Needs a lot of oomph on things that aren't present in the dataset, which is a lot of things. Also likes to throw in some weird AI version of the Magic School Bus logo, funnily enough, because that apparently was in enough of the dataset. Thankfully, flexibility isn't totally impossible with this LoRA, you just need to experiment with the prompt strengths.

Like what you see? Be sure to tip some buzz! Both blue and yellow are appreciated!

Also be sure to post what you gen with this LoRA, my LoRAs get lonely without your posts!

Also, did you know I take LoRA requests? There's a lot of things I won't make LoRAs of, but if you've got an idea, be sure to let me know!