Hi all! I hope you have fun with this one. Please share your creations here if you can, and in any case, thanks for checking it out. Cheers!
Trigger words: licking teeth, licking lips, biting lip (full prompts and ComfyUI workflows are provided in my example images)
(Trained with Diffusion-Pipe on Windows via WSL 2: https://github.com/tdrussell/diffusion-pipe/)
Description
FAQ
Comments (10)
Can it do yaeba?
Sorry, no (I actually just learned of that for the first time via your link, so thanks! TIL). Cheers!
Whats the loras tag? When I click in forgeneo nothing populates,
I'm not familiar with "forgeneo" nor the "tag" designation within that context, but if by chance these "tags" are the same as what CivitAI refers to as "Trigger Words", then for this LoRA, they're: licking teeth, licking lips, biting lip. Good luck and cheers!
@redcreamsoda tag are differnet then trigger words, It would be something like this <lora:loraname:lorastregenth>I tired the file name like usual, it doesn't seem to work
@Me007gold Ohhh, gotch'a; yeah I used those in my prompts way back in my "Automatic1111's webui" days.
That said, other than to use the structure you illustrated with the LoRA's file name as you've already tried, I'm not sure what else it could be. Sorry I'm not of much help, but good luck and take care!
sir can you do it for wan?
@fluxxes Hi there! I only have images in this dataset at the moment, and so can only train this for Wan text-to-image, but sure, I'll try to squeeze that training in sometime soon! (if/when I do, I'll tag you here with a heads up)
In the meantime, if you haven't already seen them, note that there are also a couple of preexisting Wan LoRAs for this kind of thing that you might find helpful:
Licking Lips (it can do teeth too): https://civitai.com/models/1978869/licking-lips-i2v-wan22
Biting Lip: https://civitai.com/models/2102793/luisap-wan-22-lip-biting
Cheers!
@redcreamsoda The biting lip lora for Wan is weird, in that it looks like a video playing in reverse when the lip goes into the mouth, since the mouth doesn't open, and the lip doesn't move in before biting down. But if you only have images in your dataset, it would probably have the same result.
@fluxxes @Jellai Hi there I've just uploaded a new Wan 2.1 version here: https://civitai.com/models/2205388?modelVersionId=2514512
It was only trained on images, so the intent is T2I, but it actually works somewhat for T2V too (I haven't tested I2V), and so I've included some video examples/workflows as well.
Note that, for whatever reason, I found it very difficult with this Wan version to get the "biting lip" pose working consistently. It can do it, as evidenced by the example images/video, but took me a lot of seed surfing, etc. to get there (if you happen to try it out and crack the code to getting the lip biting to work consistently, please share that information back here if you can).
Cheers!