This is my first PonyXL LORA, trained with 25 screencap images, quality might not as good. If the image generated appears to be stiff, try lower the weight or use img2img to improve results. Suggested weight (0.7~0.9).Base: towa, long hair, white hair, blue eyes, sidelocksOutfit: bare shoulders, frilled dress, frilled sleeves, long sleeves, jewelry, white dressOptional (not included in dataset but still works): (white wings:1.2)

Files with SHA256: 0425f348eccdbf8a96b917f9ebe980bcd70e42920205e190fb9e522a7d35f089