This is a LoRA that generates an image of a walk with your pet at night.

test prompt: 1girl, <lora:strollWithPetNight_20230315:0.6>, stroll_with_pet_night

test prompt: 1girl, <lora:strollWithPetNight_20230315:0.6>, stroll_with_pet_night, viewer holding leash, leash, all fours, night

It is recommended to generate a square (512x512) or a long horizontal rectangle (768x512) instead of a long vertical rectangle (512x768).

This LoRA alone works reasonably well, but if it does not work, add "viewer holding leash, leash, all fours".

(I think stable diffusion is not good with long strings like oxygen tubes or leash.)

If it is not night, please add "night" as well.

The sample images are generated by kohya-trainer/gen_img_diffusers.py by adding 0.6 to this LoRA in Anything3.1.

Since the train is NAI-based, it would be compatible with other models.