By default this will generate videos where the cum is already on the face.
I have managed to combine it with:
https://civarchive.com/models/1358184/wan-21-penis-cock-dick-lora-t2v-i2v?modelVersionId=1534254
https://civarchive.com/models/1350447/wan-cumshot?modelVersionId=1525363
Although the results are mixed. On the example video using all 3 Loras I had this one at 0.75 & the other two at 0.5.
Description
FAQ
Comments (22)
I asked for mayonnaise all over her face, not this! But thanks anyway I guess...
Doing I2V with this, every video I generate starts with a sort of yellow light flash for a few frames, almost like a camera auto adjusting to lighting levels. Anyone else have this issue? Every other LoRA I've used is fine, done lots of other kinds of generations without this happening, just this one LoRA.
yeah I tend to get the same thing on I2V with some of my loras - currently trying to figure out how to fix it. I think it seems to be a bit worse on loras trained only on still images (like this one), and ones trained on videos don't seem as bad. so far, my best solution is by making sure that the initial prompt is extremely detailed - this might not be correct but it seems like the less you're asking it to figure out, the better. I use this to put in the image which gives a very detailed description:
https://huggingface.co/spaces/fancyfeast/joy-caption-pre-alpha
having said that it might simply be something to do with the workflow or training of the lora itself - I'm still kinda figuring everything out as I go lol
I believe the model file is calling for a camera flash, so i put camera flash in the neg and it solves the problem most of the time.
It looks amazing, but I can't make it work with I2V:
I tried I2V with 1.0 it ignores the original image and morph to something like what T2V would do. I also tried 0.5 and 0.25 ...it keeps morphing and not keeping the original image and ADDS the lora to it.
This may be a silly suggestion, but are you running the T2V model? :)
I kept hitting this error and I was just literally not running the right model lmao
I did just check though, it didn't do a whole lot on my I2V checks. I think maybe cum needs to be there already?
@makiaevelio543 I only used the I2V for now, I also tried raise up the ratio to 1.2 but it just makes the face unrecognize or smudgy so I went back to 1.0 still... can't get it work, I guess it works fine on T2V (I didn't try it yet) but not for I2V so much... unless I need some settings tweak of course.
Request for Wan 1.3b model version.
is this trained for 480 or 720 Wan?
It was trained on the 14B T2V model
Can you share the prompt you used for I2V with the other Loras?
Glorious LoRA, when used at lower strength with Wan Cumshot it helps the cumshot stick
I've been trying for 3 days to get this to work in Comfy UI, using Grok for guidance on the coding, etc. Hours and hours and hours and hours. Error, after error, after error, after error, I've tried everything, I just fucking give up. I can't get it to work no matter what I do. I'm exhausted at this point.
Is it just this Lora in particular that isn't working? Can you get other Loras to work?
Grok sucks for consistency.
RIP bozo
coding??? you don't need coding. just download the required workflow and change prompt and images
You seem to have been wrong from the start.
How to use the suggestions regarding the Wan Cumshot series!! | Civitai
I've written a reference article that everyone can check out, and you can also refer to it in conjunction with my work.
how do you get it to stop changing or adding onto faces? it keeps making them rounder, or adding fake eyelashes/eyeshadow and some lipstick
we need a update for WAN2.2 I2V
it works great on 2.2. I use it on the high model only at strength 0.8
Details
Files
facials_epoch_50.safetensors
Mirrors
wan_facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
125_facials_epoch_50.safetensors
127_facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
facials_epoch_50.safetensors
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.