I can't get omnigen to work. It keeps getting stuck on processing. I have a very fast computer. Im just doing text to image without reference and it still wont work. Any help? I also randomly just get "error" on the processing screen after a while.
I noticed more frame by frame stuff on Civitai. Sadly it's mostly adult stuff, but I hope the process catches on because all low end PC or laptop owners will be able to make high quality vids as long as they can generate images based on video frames. More time and work, but that's the trade off. LOL. Awesome vid by the way.
Btw, Civitai become like a AI porn site for nerds. LOL a very special niche market porn site it is becoming now, rarely go there now, unleast I need to download AI model files.
@@TheFutureThinker Word. Most of the requests or replies I got, was for adult stuff. LOL I was just pointing out there was a way do make good vids with low VRAM but it would take more work and hopefully developers work on improving it, and they all just asked if they could make adult stuff. I was like: What about Sci Fi or action movies? Not one response. LOL.
According to my tests, this model is crap. The idea is great, but the model itself did not generate even one usable image. I already spent several hours, to get the thing going and generate tens of images. Smudges, inconsistent physics, crappy hands... I even got some gibberish text on top of one of the images... I tested with and without input images, it's the same... There was one image I was about to say I'm happy with, and then I noticed that it put too many buttons on the coat of a woman, even on places where there shouldn't be buttons.
I agree, I have been testing tons of different conditions, prompts, styles, etc... The outputs are all poor and useless. Sometimes something "okay" comes out but even then to have it actually useable would require massive amounts of manual work after the fact.
I'm following your video exactly and still can't use it because the node is still red with Import Failed. I've asked claude and others to try and figure this out, but it's still not working.
it need to be a clear image for reference to achieve that result. That why some people just randomly test it, and saw not getting result like this one.
I am not so sure why, but today is the second day I use the hugging face space for omnigen and got "error" in output. I still haven't been able to make it work locally, but still figuring out. Hopefully is a me problem, and this error in their hugging space is not related. Glad to see you talking about it! New sub
Hi, it seems to be very intersting, I would like to try it with ComfyUI, I don't understand how to load model and VAE, can you update the readme please ?
I would like to try it to, but I have comfyUI on pinokio, I think that could be a problem? With CatVTON I have problems too with having to install requirements. :(
OmniGen offers a convenient way to access a series of features that are typically available separately or through different ComfyUI workflows. It combines tools that function like Insightface, ControlNet, and inpainting, etc, into a single package, which is helpful for users who prefer not to wrestle with the complexities of ComfyUI noodles 😄 However, this convenience comes at a cost. The model's weights size of ~16 GB (partially loaded into VRAM) makes it either impractical or extremely slow on common GPUs (with 8 GB of VRAM). Additionally, to be completely honest, the functionality provided by OmniGen can be replicated using separate models outside of the package. There's almost nothing new here, as far as I can see. Nonetheless, it is a commendable effort to merge everything a regular user needs in one place.
It also costs flexibility. I often compare ComfyUI to the software Nuke which I used for many years when I worked as a digital compositor. Having things as separate nodes allows for far more flexibility than just one node that does a whole slew of things.
I wished it could. I tried it a lot now and no result are es good compared to what I did with the same picture in photoshop. Even simple stuff, like remove sky leads to a new strange image. Only the example they have in the list are working well.
the other day i was just really disappointed at the new adobe ai photo editor. i think omnigen is really a game changer. just the installation process is a bit more tedious than other nodes.
The model does not perform well in hand details, and when editing images, individual text prompts are far less effective than using images as a reference. But I think the model's consistency is excellent when it comes to image fusion.
Super slow on their site, almost useless, this is why nobody will use it in the long run because it requires too much setup and coding installations. I’ll wait for a superior model and software
I like how it can remove an object without a need of segment and masking.❤ Great vid.
Me too, just text instruction
I can't get omnigen to work. It keeps getting stuck on processing. I have a very fast computer. Im just doing text to image without reference and it still wont work. Any help? I also randomly just get "error" on the processing screen after a while.
same issue if you got solution let me know
I tried the stand alone version and it sucks arss. Tried diff promps and it could not do a basic face swap.
I noticed more frame by frame stuff on Civitai. Sadly it's mostly adult stuff, but I hope the process catches on because all low end PC or laptop owners will be able to make high quality vids as long as they can generate images based on video frames. More time and work, but that's the trade off. LOL. Awesome vid by the way.
Those img2img type video, can be done using one of the good old extension in A1111, Mov2Mov. It was my favourite before AnimateDiff.
Btw, Civitai become like a AI porn site for nerds. LOL a very special niche market porn site it is becoming now, rarely go there now, unleast I need to download AI model files.
@@TheFutureThinker Word. Most of the requests or replies I got, was for adult stuff. LOL I was just pointing out there was a way do make good vids with low VRAM but it would take more work and hopefully developers work on improving it, and they all just asked if they could make adult stuff. I was like: What about Sci Fi or action movies? Not one response. LOL.
Hi and where to put the model and why it is not in the nodes tell me
my omnigennode have latent, what shoud i do?
According to my tests, this model is crap. The idea is great, but the model itself did not generate even one usable image. I already spent several hours, to get the thing going and generate tens of images. Smudges, inconsistent physics, crappy hands... I even got some gibberish text on top of one of the images... I tested with and without input images, it's the same... There was one image I was about to say I'm happy with, and then I noticed that it put too many buttons on the coat of a woman, even on places where there shouldn't be buttons.
Nice try Adobe
@@VinciSid I've got nothing to do with Adobe. Except for despising them for their "unfriendly" customer polices.
I agree, I have been testing tons of different conditions, prompts, styles, etc... The outputs are all poor and useless. Sometimes something "okay" comes out but even then to have it actually useable would require massive amounts of manual work after the fact.
I'm following your video exactly and still can't use it because the node is still red with Import Failed. I've asked claude and others to try and figure this out, but it's still not working.
01:56 unbelievable how consistent style it can be created.
it need to be a clear image for reference to achieve that result. That why some people just randomly test it, and saw not getting result like this one.
I am not so sure why, but today is the second day I use the hugging face space for omnigen and got "error" in output.
I still haven't been able to make it work locally, but still figuring out. Hopefully is a me problem, and this error in their hugging space is not related.
Glad to see you talking about it! New sub
The architexture is really great
Can this be installed on Stable Diffusion WebUI?
Unfortunately stable diffusion web UI have not update at all
Hi, it seems to be very intersting, I would like to try it with ComfyUI, I don't understand how to load model and VAE, can you update the readme please ?
I would like to try it to, but I have comfyUI on pinokio, I think that could be a problem? With CatVTON I have problems too with having to install requirements. :(
OmniGen offers a convenient way to access a series of features that are typically available separately or through different ComfyUI workflows. It combines tools that function like Insightface, ControlNet, and inpainting, etc, into a single package, which is helpful for users who prefer not to wrestle with the complexities of ComfyUI noodles 😄
However, this convenience comes at a cost. The model's weights size of ~16 GB (partially loaded into VRAM) makes it either impractical or extremely slow on common GPUs (with 8 GB of VRAM). Additionally, to be completely honest, the functionality provided by OmniGen can be replicated using separate models outside of the package. There's almost nothing new here, as far as I can see.
Nonetheless, it is a commendable effort to merge everything a regular user needs in one place.
It also costs flexibility. I often compare ComfyUI to the software Nuke which I used for many years when I worked as a digital compositor. Having things as separate nodes allows for far more flexibility than just one node that does a whole slew of things.
Bruh ur just mad you spent hours like me on complex comfyui models that literally get out performed by a few nodes hahaha
got it working on AMD (linux). This is such a powerful yet simple node.
What are your PC Specs
I dont know, but my comfyui stuck and it wont run.
I wished it could. I tried it a lot now and no result are es good compared to what I did with the same picture in photoshop. Even simple stuff, like remove sky leads to a new strange image. Only the example they have in the list are working well.
How to run .txt in colab or how to open command in colab
Good video, but it's too slow, how to speed it up in low vram machine.
🔥
Respect you, sir. Comfyui node plz, sir.
the other day i was just really disappointed at the new adobe ai photo editor. i think omnigen is really a game changer. just the installation process is a bit more tedious than other nodes.
I tried it, without image (10 minutes), with 1 image ...1 hour 😢
It's pronounced Ah'm Nee Gen.
OmniGen - Ah'm Nee Gen
not that good ... used on image ... asked "replace the head with Image 2" ... and wasnt great ....
Seems it does a poor job of retaining the original image when it does it's 'maskless inpainting' gimmick.
The model does not perform well in hand details, and when editing images, individual text prompts are far less effective than using images as a reference. But I think the model's consistency is excellent when it comes to image fusion.
Now I am out of work
Results are often not good.
Then your prompt structure and input images need to optimize
Super slow on their site, almost useless, this is why nobody will use it in the long run because it requires too much setup and coding installations. I’ll wait for a superior model and software