Full Nelson Sex Hun | Wan Video Lora
Dataset of 76 clips with various resolutions, 1sec to 3sec, trained at 286 resolution 2e-05 Learning Rate, Num_Repeat = 3
The workflow im using is:
FramePack π img2vid / txt2vid with LoRa Workflow from @K3NK
Hunyuan π» AllInOne βAll Versions π.π±π©π§πͺ - Ultra1.5 from @LatentDream
IMG to VIDEO simple workflow WAN2.1 | GGUF | LoRA | UPSCALE | TeaCache @UmeAiRT
* you can find a slightly modded wan workflow inside the sample video's metadata
Example I2V prompt:
FU11N31S0N: A woman enjoys an anal penetration with her legs spread wide apart. The man thrusts his veiny penis deeply into her, creating a full enelson position. Her facial expressions indicate pleasure, making sounds of enjoyment. His movements are rhythmic and forceful, exerting significant effort. This intense and explicit scene is not suitable for all audiences.Example T2V prompt:
FU11N3L5ON, two individuals engage in an intimate encounter. The male lies on his back with the female sitting astride him, his hands gripping her shoulders. Her legs wrap tightly around her neck and upper torso, restricting her mobility. The male partner has his erect penis inserted deeply into her anus. Thrusts are slow and steady, allowing for gradual penetration. The female adjusts her position to stay securely atop the male. Both have fair complexions with minimal visible hairiness.Have fun! π
***** This LoRA is shared without warranties and with the condition that it is used in a lawful and responsible way. I do not support or take responsibility for illegal, harmful, or harassing uses. By downloading or using it, you accept that you are solely responsible for how it is used. *****
Description
For this to work I found that CFG has to be greater than 1.0 otherwise theres little or no thrust action....
FAQ
Comments (5)
We need a T2V version as well!
this Lora works well with other Full Nelson Lora πit doesnt seems much movement if you only use 1 lora ...
i needed to increase cfg for get movement, i havent try with the new lightv2x tho.. i might later on
After a couple of tests, I'm assuming that for I2V you have to start with an image of them already in full nelson. Is there any LoRA that can get them into full nelson?
How is the input image generated for i2v? It seems to require input to be already in full nelson position.