Version 1 - you can tweak LoRA strengths to get some different results. With strengths <code>1</code> you get exactly game model without Checkpoint style, it's hard to force different clothes etc. With LoRA strength about <code>0.6</code> the generate character get single Nightbringer Evelynn elements, sometimes tails, sometimes head gear. Version 2 - with LoRA strength <code>1</code> you get nice results, and good response for clothes.

Nightbringer Evelynn

This version contains better data set for training, so the results are better and more universal. The main trigger prompt is ```Nighbringer Evelynn```.

Files with SHA256: 2929c85f1a4361056c993b2e4c2b02ccdc12f66101b3437171ef2defac9ef2c1