This is a LoRA for creating the dreaded GIANT RAT for medieval rpgs. I'd mess around with different checkpoints to see how they interpret it. You definitely can get some variance, but it also depends on what you want. The source training images were purposefully at different distances, so try out portrait or distant shot with a very few other trigger words to see what happens on whatever checkpoint you want.

The images were made with hires.fx x2, denoising 0.23, 4x_foolhardy_Remacri upscaler

The comparatively very few trigger words for the model are the following:

animal focus, blurry, blurry background, blurry foreground, claws, depth of field, distant shot, fangs, giant rat, looking at viewer, open mouth, portrait, realistic, sharp teeth, solo, teeth, whiskers