I trained this model with art from as109 on pixiv,
This model works best at 1.0
In testing, I've found that this model has slight anatomy problems. This can be fixed by lowering the strength slightly, or using it in conjunction with other lora.
Necessary Trigger Word: as109, source_anime
These negative tags might also help: simple background, white background, emphasis lines,
Using "simple background" and "white background" as negative tends to help give the more dark/dingy color pallet that the original artist uses
For some reason, emphasis lines tend to show up every once in a while when not asked for. Use "emphasis line" as a negative if you have problems.