A bit of a problem. The natural language prompts in the trigger word sections do not appear to actually be in the tag set it was trained on when I opened the lora up. The tags this thing was trained on do not appear to be very descriptive either. The end result is that is lora is not very good at putting these outfits on characters other than Twilight Sparkle.

I would advise making a tag set to describe the outfits and write that down as trigger prompt before tagging the background and character of every image. The Danbooru page should have some good references for tagging, and you can probably get some mileage out of star print.

Edit: Also, don't tag "female" and "1girl" in the same tag. That's over tagging.