(HUNYUAN v1) Trigger word "Pokimane". Drag and drop a video for the workflow I used. Follow THIS GUIDE to train your own Hunyuan video LoRA on Windows. Love my models and want to support me? Send some buzz my way. If you're feeling extra generous and would like to donate to help support the costs of training these resource intensive models you can directly donate to me though HERE.
(FLUX v2)Likeness should be improved a lot. About half the generation have her mouth slightly open, though to be fair, that is actually how she looks a lot of time. Prompt smiling or a different expression to get rid of it. No trigger words needed.
Love my models and want to support me? Send some buzz my way. If you're feeling extra generous and would like to donate to help support the costs of training these resource intensive models you can directly donate to me though here
Description
FAQ
Comments (17)
erhm... i don't think this is a good idea after what happened time ago with those deepfakes 🤣 but whatever.. 🤣
I am sure there is one editor who's really gonna love this tho XD
T-Swift sued for that stuff and they still trained her as a Video Model :)
This resource is intended to reproduce the likeness of a real person. Out of respect for this individual and in accordance with our Content Rules, only work-safe images and non-commercial use is permitted.
If you are this person or a legal representative of this person, and would like to request the removal of this resource, you can do so here.
That cat is out of the bag, Pandora's box lay in smithereens, ain't no horse in that barn, etc. Trying to tamp down fakes from here-on-in would be an eternal game of whack-a-mole where they'd need more arms than Ganesh and Goro combined.
I would assume its safe to say you will do a valkyrae one as well? :D
Correct. I have all my datasets from FLUX models. So if you're curious which models I will be uploading, have a look at my FLUX models. 😁
@AIENGI Can't wait for Mia Khalifa
your loras are amazing for hunyuan compatible with multiple loras please keep up the good work hope to see more celebrity loras from you ;)
More coming soon. Kyedae is training. Alex Botez next then some other twitch streamers. Check my profile. If I did a FLUX model of them then 99% sure I will do with for Hunyuan. Cheers.
I think there's something wrong about the playback speed of your videos, check if your frame rate is higher than 24 on the video combine node
I run into this issue a lot too with the correct frames on the video combiner. I don't know exactly what causes it but i think it's a mixture of other settings. sometimes takes playing around with guidance and flow shift and strength until right combination is found is my theory. I am going to try generating some for this lora to see if i have same issue or not
😍 👍🏻
I have issue when i'm downloading your video (in different way) then i drag to comfyui. It keep saying there is no workflow. Someone has solution ?
you ought to be able to see the prompt and details if you look at the file properties -> details -> comments
if they are available at all
i am also having trouble using the workflow, when i drag the video into comfyui there is an error saying a bunch of nodes are missing. any ideas?
@bj4201000 You'll need to download the missing nodes that are in red. However I would like to know, how did you get the workflow working? The drag and drop ain't working for me.