<h2 id="heading-11">Laure Valée (Bulii)</h2>First version is called V3 to match up with the file name, so there is no V1 or V2.V3 was accidentally trained without any image captioning but still provided good results. V4 is re-trained with almost identical settings but with image captioning, but actually gives me worse results. Putting both up to give you the choice of whichever works best for you.All images in showreel are simple generations except the anime one which also has highres fix.As with all models of real people, don't be weird.

Laure Valée - LaureBV - [LoRA]

I accidentally forgot to include captions in my V3 training, but I actually get better results with it. Feel free to experiment with either one and see which gives you better results.

Files with SHA256: 63ce2280490433cca1775f3f2577c0bf706349eadaf17fc93349203d836040a5