This LoRA was trained with more than 300 images obtained mostly from: https://fancaps.net/

The recommended weight is 0.80-0.90 and the Sampler, Euler a, but it should work fine with others.

You can get closer to some characters in the anime, but the descriptions must be very specific.

It is recommended to use HiRes.fix and Adetailer for full body shots.

Anyway, I hope you like this LoRA and if so, please leave me your feedback.