The "how" is important but the "why" is the best part of your videos. I've corrupted my system and goofed up things up many times. This really helps. Thanks!
loved the video, but I was able to get it installed and used it last night. But this morning I have no idea how to start omnigen up again. In roop I just click on the bat file. How do I start onmnigen once it is working properly. Thanks
thanks a ton the issue was i was using CUDA 12.6 with Python 3.12 which was a total mess and kept throwing errors. now it's all good and working your the best 🔥
One question, if you use a virtual environment, that means you need to install again Omnigen every time you want to use it? and Cudatoolkit and everything else? As other people said, it´s great that you explain everything so we learn and understand what we are doing, thank you for that.
@@foreropa so how a virtual environment works is that the dependencies are stored in a directory on your computer. That means that every time you use it you call it and effectively turn it on. For instance using python you would use a command that looks like call venv/scripts\activate. Orwith conda you would use conda activate myvenv .. What you then do is create a batch script that activates it and then runs the program. Or in other words a starter file
@@cognibuild Thank you for the quick answer. I´m sorry, but I´m new at this. So to understand correctly, when you want to use it again, you should follow again your instructions and it just activates what is already installed? So it would be much faster than the first time, I assume?
@@foreropa correct.. you create a "script" that does multiple commands. If you get the omnigen script from the patreon site it has a "run" file already built but you can also create it yourself. I have another video on bat files
I got this error: ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. omnigen 1.0.3 requires torch
@@cognibuild Yes. R4060 6GB VRAM. I installed CogVideoX successfully yesterday. It seems when I pip install pytorch, it installs torch 2.5.1, so maybe if I specify pytorch=2.4.1, it might work. I'll try that. Is torch the same as pytorch?
It looks like when you pip install torch, it takes the latest version 2.5.1, so I replaced torch with torch==2.4.1 and it seems to work. Apparently OmniGen 1.03 doesn't work with torch 2.5 and above.
I'm merging 2 pics like you did for a bar pic and its been going for over 2 hours and not finished. Is this normal? I have a very hi end PC gaming rig with Nvidia card and plenty of ram.
@@alanlovern4147 no that's not normal at all. I'm going to bet you don't have your cuda installed correctly. Come find me on discord and I'll see if I can help sort you out
Python was not found; run without arguments to install from the Microsoft Store, or disable this shortcut from Settings > Manage App Execution Aliases. what will i do?
@ComUser-x2d you need to set up miniconda. Go to my videos and check out AI setup. You need to get your system set up . If you want an easy way then go to my patreon site and get the system installer. It will do everything for an NVIDIA system: www.patreon.com/posts/117205365
i have 2 questions.. 1- what is the path pf the venv that download all the stuff in it? 2- can i download the big 15gb file from a direct link and just put it in its path? because i have slow internet and the cmd sometimes cutff start over ..
@@QorQar you have 16gb vram? ok.. irst of all you dont need cuda 12.6.. i would go back to 11.8 .. also make sure you have the rest of the needed installs. Here is a video.. make certain you all these installed: th-cam.com/video/b0kPIQi-4N8/w-d-xo.html
omnigen is really deceptive, it seems that it was overtrained with their demo visuals,who works perfectly,but when you injects other visuals inside its really not so accurate...
![Shoot EPIC MOVIES with this FREE AI Tool! [ComfyUI Tutorial + Free Workflow]](/img/n.gif)
The "how" is important but the "why" is the best part of your videos. I've corrupted my system and goofed up things up many times. This really helps. Thanks!
that was THE best install guide ever, super BIG thanks!
Great video, love the casual style and attention to details.
loved the video, but I was able to get it installed and used it last night. But this morning I have no idea how to start omnigen up again. In roop I just click on the bat file. How do I start onmnigen once it is working properly. Thanks
I just get the error: "ERROR: Could not find a version that satisfies the requirement torch
come over to Discord and lets sort it out.. add a post to Troubleshoot:
discord.gg/swPN58J22U
Wonderful guide , I want to know everytime i want to use it i have to wait that much time for the package to load again ??
Thanks, great tutorial as always! installed this app for myself two days ago, it's really interesting and handy for some tasks.
i have 3060 6gb and it is taking 20 minute to generate a simple prompt :( i wish i had 24gb graphic card :(
@@vs3228 the struggle is real
thanks a ton the issue was i was using CUDA 12.6 with Python 3.12 which was a total mess and kept throwing errors. now it's all good and working your the best 🔥
glaf you got is sorted out! Onward and forwrd!
I have this problem and I don't know how to solve it. I would appreciate it if you could guide me. Regards.
@@jbricenio which problem?
One question, if you use a virtual environment, that means you need to install again Omnigen every time you want to use it? and Cudatoolkit and everything else? As other people said, it´s great that you explain everything so we learn and understand what we are doing, thank you for that.
@@foreropa so how a virtual environment works is that the dependencies are stored in a directory on your computer. That means that every time you use it you call it and effectively turn it on. For instance using python you would use a command that looks like call venv/scripts\activate. Orwith conda you would use conda activate myvenv ..
What you then do is create a batch script that activates it and then runs the program. Or in other words a starter file
@@foreropa search my videos and I have one I made some months ago that explain virtual environments in greater detail
@@cognibuild Thank you for the quick answer. I´m sorry, but I´m new at this. So to understand correctly, when you want to use it again, you should follow again your instructions and it just activates what is already installed? So it would be much faster than the first time, I assume?
@@foreropa correct.. you create a "script" that does multiple commands. If you get the omnigen script from the patreon site it has a "run" file already built but you can also create it yourself. I have another video on bat files
@@cognibuild Oh, ok, great to know it, thank you so much!! I subcribed to your page, thanks a lot for all your help.
Well, I figured it out. Love the video. I
I got this error: ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
omnigen 1.0.3 requires torch
@@leeweichoong do you have an Nvidia card?
@@cognibuild Yes. R4060 6GB VRAM. I installed CogVideoX successfully yesterday. It seems when I pip install pytorch, it installs torch 2.5.1, so maybe if I specify pytorch=2.4.1, it might work. I'll try that. Is torch the same as pytorch?
It looks like when you pip install torch, it takes the latest version 2.5.1, so I replaced torch with torch==2.4.1 and it seems to work. Apparently OmniGen 1.03 doesn't work with torch 2.5 and above.
I'm merging 2 pics like you did for a bar pic and its been going for over 2 hours and not finished. Is this normal? I have a very hi end PC gaming rig with Nvidia card and plenty of ram.
@@alanlovern4147 no that's not normal at all. I'm going to bet you don't have your cuda installed correctly. Come find me on discord and I'll see if I can help sort you out
How do i install on macos git ?
Python was not found; run without arguments to install from the Microsoft Store, or disable this shortcut from Settings > Manage App Execution Aliases. what will i do?
@@ComUser-x2d you need to install python. Go search my videos and you'll find a step-by-step installation tutorial
@@cognibuild 'conda' is not recognized as an internal or external command,
operable program or batch file.got this error what to do then?
@ComUser-x2d you need to set up miniconda. Go to my videos and check out AI setup. You need to get your system set up .
If you want an easy way then go to my patreon site and get the system installer. It will do everything for an NVIDIA system:
www.patreon.com/posts/117205365
may i know your system configuration
@@RAMAMOORTHI.A I have a RTX 4090, pro art x670e motherboard, amd ryzen 9 CPU , 128 GB ram, 1& 2 tb SSD
i have 2 questions..
1- what is the path pf the venv that download all the stuff in it?
2- can i download the big 15gb file from a direct link and just put it in its path?
because i have slow internet and the cmd sometimes cutff start over ..
1. the venv folder is found in ./Omnigen/venv
2. Im uncertain abuot how you would do this.. the software just downloads it into cache as needed
omg that helps a lot. Thank you
with
i7
32ram
16vigs
Not run
and not I did not see the Vega despite downloading Coda 12.6
@@QorQar you have 16gb vram? ok.. irst of all you dont need cuda 12.6.. i would go back to 11.8 .. also make sure you have the rest of the needed installs. Here is a video.. make certain you all these installed:
th-cam.com/video/b0kPIQi-4N8/w-d-xo.html
bro i love you like i love trump bro you delete the bs in your efforts to show how it all goes down man! thank god for you! ima do your Patreon too
omnigen is really deceptive, it seems that it was overtrained with their demo visuals,who works perfectly,but when you injects other visuals inside its really not so accurate...
lov
Please Co-ordinate the sound of your voice with mouth movements. Very irritating when sound is later than motion of lips.
Yes I was having problems with my system at that point. Hopefully things have sorted out. Thank you for hanging in there