This lora was trained on the I2V model using 25 high quality videos from different angles (behind, side, front and POV), for all your (consensual) spanking needs. Should be flexible enough that as long as a butt is in frame, the lora will work with both hands and paddles.
Trained using diffusion-pipe for 3200 steps, 128 epochs. Learning rate of 3e-5.
Try the following prompts to control the action:
A {man|woman} {appears from off screen and|} spanks her buttocks {multiple times|once} with {their hand|a rectangular paddle}, {using big and fast arm movements|gently}.Workflow and full prompts are attached to the videos.
Based on other Wan loras I've trained it should still work with T2V, but untested.
Description
FAQ
Comments (28)
Nothing beats a little pat of encouragement
T2V planned?
I was commissioned to train this one I2V, so no plans to but people have got good T2V results out of my Masturbation Cumshot lora which was also trained I2V, so I'd say give it a try if you haven't.
Titty Slap plz
Kind of works, alongside other Loras - Check out the videos I posted.
@aipinups69 Oh nice, I thought it might, but thanks for testing!
Alright, finally, a decent spanking lora!
Well, the physics are real.
consider to add it to tensor.art
Could somebody share the workflow you use for the generation? or at least key parameters set up? (steps, shift, frame rate)? whatever I try the hands movement is too fast....not sure how to slow it down :-(
How to import the workflow from the examples? im trying to drag and drop but nothing happens
Hey I just checked and I can still drag these into Comfy after I download them for the workflow. Is your Comfy up to date?
@definitelynotadog icannot even drag them, i try to download and rename as wbmv and drag but nothing happens me confy its the latest version
I'm new to this but when I use this lora it seems to work well but I also end up with what appears to be image's from the training data overlayed on the figure. Like a bruised ass or hairy ass. Is it possible to just get the physics and not the visuals? Is it possible to just have a lora that makes it so when I write "figure one goes over figure 2's lap and get's spanked" wan will know what to do and figure it out?
I have seen the same thing, it also seems to affect the overall contrast of the video, degrading the quality somewhat, which is unfortunate, because the physics are great.
what resolution are you rendering your image at? the smaller the resolution, the better the results, generally.
I don't know EXACTLY what you are talking about, but if you are talking about "Ghosting", that can happen when the ai is confused about "where to go next". Ambiguity in the prompt can create "parallell universes" where the video takes different directions that are layed on top of one another.
Why is it that in all the videos I come out in, it's always about touching the buttocks instead of slapping or hitting them?
Perhaps it was because I was using Wan2.2, but after I reset the weight to 3, the effect of slapping the buttocks with hands was achieved.
This are working just fine on Wan 2.2
I used this workflow
https://civitai.com/models/1824577?modelVersionId=2068437
is somebody going to improve this further by training it with wan 2.2?
Anyone else get this error on the Gradio app?
Error loading LoRA LoRAs\spanking_for_wan_v1_e128.safetensors: only integer tensors of a single element can be converted to an indexPlease do a wan2.2 version. It doesn't really work "fine" in wan2.2 as some say. At least not when it comes to spanking with a paddle. In Wan2.1, hits in general suffer from a strange impact particle effect. Wan2.2 doesn't have the same issue but the lora doesn't work well on it. The paddle is used more like a hammer which is exactly what happens in wan2.1 too without this lora.
wan2.2 i2v? any plans?
I think the name is wrong, it should be called Parkinson Lora
I think the name is wrong, it should be called Parkinson Lora
2.2 pls sir.
anyone going to do this for ltx2. would be awesome with some sound
this person doesnt respond to anyones comments
Details
Files
spanking_for_wan_v1_e128.safetensors
Mirrors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128 (2).safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
spanking_for_wan_v1_e128.safetensors
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.