I've created a Lora because there's a surprising lack of simple scenes depicting people drowning, and Stable diffusion's 'drowning' doesn't function as desired. The Lora is primarily based on images where only the upper body is emerging in a sea with waves. However, during the operational check of the created Lora, I found that including prompts related to the lower body, like skirts or socks, results in images of standing by the sea rather than being in it. Hence, please exclude prompts related to the lower half of the body. On the other hand, for one-piece costumes like in Tokyo Mew Mew, increasing the Lora's strength makes it easier to generate images with the lower body submerged in water. Conversely, reducing the strength will result in images of wandering in shallow waters rather than drowning.

Files with SHA256: 33ae9f3675fc68ddc098767c205234ab611e2d4d13708286539576b2bf9a0ef8