This is an experimental LoRA I'm developing, and when it works it works like this: You feed it an image and it will take that person's likeness and environment and create multiple scenes inside a 97 frame window (this LoRA was specifically trained for 97 frames). This first version should create three quick scenes from your input image: 1 - The person undressing. 2 - The person posing completely naked. 3 - The person sitting and spreading their legs. The fps is slightly sped up so you should run it through interpolation to get it back to a more natural speed.

The prompt you use should look something like this: The same woman starts to take off her clothes, cut to the next scene where the same woman is completely naked and posing nude, the scene then cuts to the same woman opening her legs to reveal her genitals with her vagina exposed.

Use the Lightv2x lora with this, so that you can run at 4 Steps, CFG = 1
<a href="https://huggingface.co/Kijai/WanVideo_comfy/blob/main/Wan21_T2V_14B_lightx2v_cfg_step_distill_lora_rank32.safetensors">https://huggingface.co/Kijai/WanVideo_comfy/blob/main/Wan21_T2V_14B_lightx2v_cfg_step_distill_lora_rank32.safetensors</a>

The training was done at a 2:3 aspect ratio, so you can use these resolutions:
256x384, 384x576, 512x768, 640x960, 768x1152, 896x1344, 1024x1536

<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/e779d83f-e80e-4f3f-b69a-c7ed392be618/width=525/e779d83f-e80e-4f3f-b69a-c7ed392be618.jpeg" /> 

<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/82fcae1d-0419-4d9c-a4de-ff57b6729698/width=525/82fcae1d-0419-4d9c-a4de-ff57b6729698.jpeg" /> 

<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/72a94000-b3c5-47bc-8c71-370b9382a788/width=525/72a94000-b3c5-47bc-8c71-370b9382a788.jpeg" />

Files with SHA256: 65300834e2e51082a1b0e26795ee16d44a516a22e25bbdfcbd7b868aaf6bde09