*salutes*

Works great, if a bit overfit to show's style and its artifacts (expected for 200 image LoRAs and to some degree desired). Some tips
1: You can cut out "human" and "female" from the tags. Human is default, and 1girl handles identifying her as female just fine. In-fact you can cut most tags about body shape (there's a few that auto-label seems to have added) and models will learn them as just part of the character (you do want to include those for styles or if they don't match the rest of the data.).
2: You seem to have applied a lot of redundant and non-tags ("1980s cartoon style" should be "1980s (style)": When training you're trying to tell the model what's in the image so it can compare and contrast to things it already knows, so not using standard tags is at best not efficient) in tagging. I absolutely wouldn't tag most pics of this show, let alone all of them, best quality and high detail. In-fact, you don't need to tag style identifiers in character unless you have multiple styles you deliberately want to separate.
3: Speaking of which, a technique I haven't fully tested for this purpose (but what I have done with it suggests its possible and it absolutely works for controlling other bits of style) is that you can selectively label screenshots from the show with an existing tag applicable about their quality ("low res", "vhs artifacts", or "jpeg artifacts") and include images from sources outside the show proper (I know there's an behind the scenes book for the show out there with lots of art that's still in style) that doesn't have those artifacts (and thus not tag) and it will learn the artifacts are artifacts.

Disable visibility of X and XXX images if you can’t see your model but can see the notification for this comment. Credit card companies have mandated this model be hidden to most users.