Trained with 69 selected images with Dreambooth LORA based on "Anything-v4.5" model, might be some hit and miss due to limited resources.
Recommended weight around (0.5~0.8), if the pictures appears to be too "stiff", try lower the weight more.
Ex. Prompt: 1girl, a girl named pursena, pursena, large breasts, swept bangs, red eyes, dog ears, animal ears, tail, gray hair, long hair, uniform, long sleeves, white jacket, red pleated skirt, red bow