Stylization by illustrator Boris Vallejo.

244 images were used to train the model.

the training of the LoRA was on the RevAnimated model.

Training with a resolution of 512x512 images

I recommend using ControlNet, without it, most often, the poses are very strange)