Help clone an I2V space without a filter
There's an I2V space that can do NSFW perfectly, but they recently added a filter to it. Can I ask you to clone it, but remove the filter?
sure you can do anything 😃
sure you can do anything 😃
No, I wasn't asking to clone your space, I was asking to you to clone another space with a nsfw filter and remove the filter. I can't do it because I don't have a pro acccount. Would you be willing to do it? I can help where I can.
sure you can do anything 😃
I can't, I don't have a pro account, that's the issue. Would you be willing to help me? 🙏
same too i don't have a pro account, but i got free 2 sapce quota? You didn't get it?
sure you can do anything 😃
I can't, I don't have a pro account, that's the issue. Would you be willing to help me? 🙏
same too i don't have a pro account, but i got free 2 sapce quota? You didn't get it?
Nope, seems like they removed it. Maybe you can put it instead of one of your spaces?
Have you ever created a free space that uses ZeroGPU before?
Have you ever created a free space that uses ZeroGPU before?
Nope, it seems like when I registered they already removed it
hey is not that mean
this i have alreay 2 space now if i'm delete some spaces i can get back to tey new model
i try delete thise for you man
now i have only one
You will see a buy promotion notification, but it's gone.
you just add token and run
I hope you will understand 😁
I hope you will understand 😁
What?
The space that needs to be duplicated is https://huggingface.co/spaces/FrameAI4687/Omni-Video-Factory/tree/main, can you duplicate it?
yes sir for you i do it now 😉
The space that needs to be duplicated is https://huggingface.co/spaces/FrameAI4687/Omni-Video-Factory/tree/main, can you duplicate it?
yes sir for you i do it now 😉
Oh really, you did, thank you! But I wanted you to remove the nsfw filter. Can you do it? I can't code, but I have an AI helper's advice on how to. Would you be willing to work with me to remove the filter?
puase filter now
Also it has a "models are not ready
ZeroGPU elapsed: 0.0s"
message
noth1ngburger
5 minutes ago
Also it has a "models are not ready
ZeroGPU elapsed: 0.0s"
message
See translation
wait
I'm sorry, but I don't know if I can fix it anytime soon because it's very complicated. but i'm trying
I understand it's not easy. I had an AI build the package for me, can you deploy it? Can't send it here though, is there anywhere I can contact you in DMs?
"To explain the technical side: the root problem is still the model files. The original creator didn't include the 'Model Weights' in the public repository. Instead, they fetch them using a hidden private key (OMNI_VIDEOS).
Because of this, even if we completely rewrite the code and successfully bypass the filter, this cloned project still won't be able to generate any videos since it lacks the actual model to process them. To run this exact code, we would need the original model weights installed in the system sever owner space.
There is only one real workaround: we have to grab a different model and mimic the workflow and small aspect ratios of this one to save ZeroGPU quota. However, the code controlling this space is around 3,500 lines long. That is a massive job, and it's definitely not going to be finished right now or anytime soon.
I do have a plan in mind, though. We can swap out the core model while referencing the structure and aspect ratios of this one. I believe using WAN2.2 AOT could do the trick. It would significantly reduce ZeroGPU consumption and wouldn't have any filters.
Give me about 3 to 5 days to work on this, though I can't promise exactly how long it will take." 😟
"To explain the technical side: the root problem is still the model files. The original creator didn't include the 'Model Weights' in the public repository. Instead, they fetch them using a hidden private key (OMNI_VIDEOS).
Because of this, even if we completely rewrite the code and successfully bypass the filter, this cloned project still won't be able to generate any videos since it lacks the actual model to process them. To run this exact code, we would need the original model weights installed in the system sever owner space.
There is only one real workaround: we have to grab a different model and mimic the workflow and small aspect ratios of this one to save ZeroGPU quota. However, the code controlling this space is around 3,500 lines long. That is a massive job, and it's definitely not going to be finished right now or anytime soon.
I do have a plan in mind, though. We can swap out the core model while referencing the structure and aspect ratios of this one. I believe using WAN2.2 AOT could do the trick. It would significantly reduce ZeroGPU consumption and wouldn't have any filters.
Give me about 3 to 5 days to work on this, though I can't promise exactly how long it will take." 😟
The AI helper I'm using decompiled their code and mimicked their workflow, we only need to deploy it now, I need to send you the files, can I private message you somewhere
Anyway here are the files: [private link], just put them in a space, no need to do anything else yet. Better create a new one for that I guess
I received the file. You can hide the link now. Thank you very much. and also If you want NSFW content, you can use WAN 2.2 for now because LORA controls NSFW in my space But you need to lower the image resolution Advanced setting quality at 4 or 5 , otherwise it will consume too much ZEROGPU time..4 step and try to gen
I received the file. You can hide the link now. Thank you very much. and also If you want NSFW content, you can use WAN 2.2 for now because LORA controls NSFW in my space But you need to lower the image resolution Advanced setting quality at 4 or 5 , otherwise it will consume too much ZEROGPU time..4 step and try to gen
I tried to use your space, it doesn't act the same. Will you put up my files on a space now?
Yes, everything was uploaded at once, but I can't guarantee it will work. However, I have fulfilled your request by uploading all the files you provided.
Have you deleted all the original files and put mine instead?
Looks like not, you better start a new space and put all the new files from a fresh start and follow the instructions
Could you possibly give me access to your account to try it myself, it would be quicker 😅
You need to start clear, delete the Omni Factory copy altogether and put my files on a new space, nothing more
It works!!!! But whose account is that?
Could you possibly give me access to your account to try it myself, it would be quicker 😅
😂🫵
Could you possibly give me access to your account to try it myself, it would be quicker 😅
😂🫵
Lol, I know how that sounded, sorry, but in fact, it's more convenient
sorry i can't give you my ID it my gmail i'm sorry
Could you possibly give me access to your account to try it myself, it would be quicker 😅
😂🫵
Lol, I know how that sounded, sorry, but in fact, it's more convenient
sorry i can't give you my ID it my gmail i'm sorry
Can you create a separate space to test if an a bit different version would be better? I'll send you the new files once again
Could you possibly give me access to your account to try it myself, it would be quicker 😅
😂🫵
Lol, I know how that sounded, sorry, but in fact, it's more convenient
Dude, there would be nothing to steal even if I wanted to, it's not even a paid account 😁
The files for an edited version: https://mega.nz/file/0QgiSLbK#qj4D_pzqU67iD9AFYwLRIQcL8wlJtysUlSkV_UQd13Y
Hey, your Frame Interpolation makes videos shorter instead of smoother