Requested LoRA

SDXL LoRA trained with Civit-AI trainer for Pony V6/Pony merges, an experimental one trained on a certain body-type of dragon. A stocky, bottom heavy type often seen in cartoons and pokemon. Such as Charizard, Goodra, Lugia or ones seen in the dragon forms of those such as Malecifant, Puff the magic dragon and the like. When making this LoRA, I put together dragon pictures from across multiple artists. Both pokemon and original characters. Though the model might produced some mixed results and require some testing in strengths.

Uses E621 tagging, no trigger words. Though having ‘pear-shaped figure’, ‘wide hips’ and ‘slightly chubby’ n the prompt can have it lean to what it’s supposed to do.

A starting prompt string for use of this model might be:

<lora:Pear-shaped_Dragons_Concept_PonyXL:1> wide hips, pear-shaped figure, slightly chubby, western dragon, score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_furry, source_cartoon,