Hellew sir could you pls show me full material node .. of this vortex effect
@basarozbent47402 ай бұрын
Hey Paul, finally got openvino plugins up and running, sadly I should say the video now is out of date and does not follow the installation process. The process in command prompt now never asks the models cause you get to load them in gimp. I have watched your video in the previous installation but later it was misleading. Good news is, SD in gimp is better and easier to use
@basarozbent47402 ай бұрын
Long story short, installation works as intended but it never installs model settings and produce crappy low res images
@YalovainkYalova3 ай бұрын
I think I did everything right, but before your video, I followed the instructions of another video... it didn't work. Then I tried according to your video. It didn't happen again. I don't know why. I followed all the instructions before you. The model upload link on hugging face in the model upload section has been cancelled. Then I found your video. I followed everything you said. Sorry for my failure. There is a problem. Do you think we can install it together if I connect remotely? thanks. If you help me, I would like to present you a Turkish rap. :)
@antiWhiteism7775 ай бұрын
It's like U-Force for the NES.
@Journey_to_who_knows5 ай бұрын
I don’t understand what you’re supposed to do, those instructions don’t tell anything, python is installed and now what, how do you get web ui open. It just throws me a bunch of errors after I used first time run and try to open any of the web ui files
@Bronco5417 ай бұрын
so what I gather is AI tools such as this is not yet ready for non-coders.
@Rosechuchan7 ай бұрын
I see why you need to use AI
@M.Vriesde-y7i7 ай бұрын
Hi. Thanks for these efforts. Unfortunately, I am stuck. Up to 5:19 of the video, everything seems OK. Software and model are installed, and the plugin path is added to preferences plugin folders (C:\Users\Public\GIMP\gimpenv3\Lib\site-packages\gimpopenvino\plugins), and I restarted. However, no OpenVino plugins whatsoever are shown in Gimps' layers menu, so Gimp apparently refuses to integrate them. Versions: python 3.12.3; gimp 2.10.36. If anyone has an idea how to proceed, I'd be thankful.
@BobDuffy7 ай бұрын
Sorry about that. The video is old and the repo has been updated. I suggest following the repos instructions more than my video. I will share your issue with the developer. Are you using the experimental version of GIMP?
@M.Vriesde-y7i7 ай бұрын
@@BobDuffy I tried, but no success. (latest stable GIMP). Anyway, thanks, I'll look for different software. Cheers.
@basarozbent47402 ай бұрын
hey there you should work with gimp 2.99 for these developing functions to work. That is an unstable version where some features are tried and tested for the more advanced users.. but I cannot promise it would work as planned🤪, many people have a hard time getting openvino up and running, including myself
@M.Vriesde-y7i2 ай бұрын
@@basarozbent4740 Thanks for the tip.
@eddyjuillerat8358 ай бұрын
WHat about denoising passes?
@BobDuffy8 ай бұрын
This is an old video. Denoising at this time was done using the compositor. And via the compositor my understanding is you could use the node system to denoise multiple passes. Now with AI Denoising today its done mostly using the Render Properties. I suppose you could still use the compositor and denoise specific render passes, but I dont believe its necessary
@venomsoffice559610 ай бұрын
Bob only one guide to GIMP with stable diffusion using Intel's Ultra Processor and using Npu for image generation .... I saw it working at the Intel Ultra Series Launch but its not working now
@BobDuffy10 ай бұрын
Sorry about that. I have not tried it recently, with NPU. Ill take a look
@galaxysquad801510 ай бұрын
Amazing how would you make an object get sucked into the vortex?
@MiDaY_i5 ай бұрын
JUST GET HOUDINI AT THAT POINT LMFAO
@carloscervantes83610 ай бұрын
Nice video Bob, I was able to set everything up and run it with no issues on a Bifrost A770 16GB. This is exactly the type of content that makes me look forward to Battlemage.
@basarozbent474010 ай бұрын
hi there, great walkthrough. I think I followed written directions on github but that really helped me figure out why I'm doing what I'm doing. I was kinda down when I realized my iris xe was no good for image generation but completing the installation and Bob's excitement(even though he is not doing it for the first time) as he builds the image really put a great smile on my face. I believe running the bat files from the right directory matters, failed a few times, I'm not really an expert, but It works on this day.. Cheers everyone
@basarozbent47402 ай бұрын
today my gimp stopped running openvino plugins due to a gdbus.exe error... trying to reinstall it and keep failing at it! I'm afraid it really is a pain to properly setup and get running these plugins
@battleforevermore10 ай бұрын
Could you do a follow up video about this? An update as well as how this has helped you overall?
@BobDuffy10 ай бұрын
I will be re-reviewing the repo instructions and plugin features. Not sure if video. Might be an article. Ill link here.
@duffyrides11 ай бұрын
Hey Bobb, I have the new Intel Core Ultra i7 155H in MSI Prestige 16 and am testing this out! Do you know if the plugin is optimized for the new integrated ARC GPU?
@IndraYarka11 ай бұрын
hi helpplease with image prompt @BobDuffy Traceback (most recent call last): File "C:\Users\ANTRA\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\_ops.py", line 565, in __getattr__ op, overload_names = torch._C._jit_get_operation(qualified_op_name) RuntimeError: No such operator ipex_prepack::convolution_prepack File "C:\Users\ANTRA\AppData\Local\Programs\Python\Python310\lib\site-packages\intel_extension_for_pytorch n\utils\_parameter_wrapper.py", line 500, in conv_prepack self.op_ctx = torch.ops.ipex_prepack.convolution_prepack( File "C:\Users\ANTRA\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\_ops.py", line 569, in __getattr__ raise AttributeError( AttributeError: '_OpNamespace' 'ipex_prepack' object has no attribute 'convolution_prepack' Total time: 15.75 seconds and open discord may be
@TheWayneStevenson11 ай бұрын
Bob, first off all thanks for sharing your articles, and videos. I just tried sending a link to someone on Reddit looking to use his Arc a770 with Stable Diffusion and I just went through this a couple weeks ago and your article was a huge help. Wouldn't have had such an easy time without you. Thanks again. At any rate, your Intel articles are all 404ing on the Intel site. Specifically looking for "Three Ways to Generate AI Art Using Intel Arc GPUs" as I know it would help him out. Do you have these articles published anywhere else that I could share them with him?
@IndraYarka11 ай бұрын
input image image prompt AttributeError: '_OpNamespace' 'ipex_prepack' object has no attribute 'convolution_prepack
@BobDuffy11 ай бұрын
Did you follow instructions on Discord? See my link in the other comment
@IndraYarka11 ай бұрын
@@BobDuffy hi yes of course but you have mistake in instructions
@IndraYarka11 ай бұрын
@@BobDuffy input image image prompt AttributeError: '_OpNamespace' 'ipex_prepack' object has no attribute 'convolution_prepack not work from me
@Ray88G11 ай бұрын
Thank you for doing these videos . I hope more people were interested in Intel GPU's because they kick asss
@Ray88G11 ай бұрын
Very exciting . Really looking forwards to Battlemage
@jorinryms294311 ай бұрын
For anyone watching, the instructions on the GitHub repository do not work. It "requires" Python3.12 but certain dependencies don't exist for this version of Python. This is stated, so I have been trying with 3.10 for the last couple of days. I would either make corrections to the batch script or manually plug commands into powershell. Now, I at least have the option to download models. Note: you actually don't need to rename the openvino folder unless it is in the home folder. Apologies if I missed that in the tutorial. Very annoyed at the moment, but I figured it out. In order to work correctly, `install.bat` needs to be edited to match the specific Python version you have on your system. If anyone else is experiencing issues, perhaps I will fork the repository with fixes available. What works for me: Windows 11 Python3.10 GIMP 2.10 openvino (via pip) 2023.2.0
@BobDuffy11 ай бұрын
Sending to dev team for them to check out
@pixelbat Жыл бұрын
Is Fooocus support for arc new? Does it require modifying the install?
@BobDuffy Жыл бұрын
Fooocus can use Arc with IPEX similar to how SD.Next and Comfy do but its a custom solution not yet documented. Hopefully we can publish an article and point to a repo on this soon as we did with GIMP, A1111 and SDNext
@frankjust394611 ай бұрын
@@BobDuffyI am also interested, can you please let us know when is it gonna be ready? Thank you so much
@BobDuffy11 ай бұрын
@frankjust3946 will do. Next week I meet with the owner of the forked repo which has support for Arc on Windows, and will collaborate to create installation instructions
@BobDuffy11 ай бұрын
FYI @pixelbat and @frankjust3946 I added Fooocus installation instruction the the GenAI channel on the Intel Insiders Discord. Go here to check it out discord.com/channels/554824368740630529/1175245000322322483/1175245000322322483
@IndraYarka11 ай бұрын
@@BobDuffy File "C:\Users\ANTRA\AppData\Local\Programs\Python\Python310\lib\site-packages\intel_extension_for_pytorch n\utils\_parameter_wrapper.py", line 500, in conv_prepack self.op_ctx = torch.ops.ipex_prepack.convolution_prepack( File "C:\Users\ANTRA\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\_ops.py", line 569, in __getattr__ raise AttributeError( AttributeError: '_OpNamespace' 'ipex_prepack' object has no attribute 'convolution_prepack' Total time: 12.09 seconds
@slovek Жыл бұрын
I followed all of the steps and this did not work. Used the correct versions etc.
@BobDuffy Жыл бұрын
It appears the latest version no longer supports a simpler installer described in this video. Did you follow the instructions on the GitHub page which requires you install OpenVINO?
@slovek Жыл бұрын
@@BobDuffy I did. I thought I had messed up a step and redid all steps just to be certain but it did not work for me.
@BobDuffy Жыл бұрын
@@slovek what error do you get. Can you send me a screenshot of the CMD window
@BobDuffy11 ай бұрын
Are you using the experimental version of GIMP 2.99.14 and either Python 3.9 or 3.10?
@theftking Жыл бұрын
Make OIDN temporal when rendering animations, please.
@kokoze Жыл бұрын
Oooh that looks good!! :)
@ronihalabi Жыл бұрын
better than rtx?
@PrefoX Жыл бұрын
RTX is just for games, you mean CUDA/optix and Nvidia is faster, but more expensive.
@ronihalabi Жыл бұрын
@@PrefoX yes that's what i meant i was cooked lol thank you
@rano12321 Жыл бұрын
Odin keeps more details than optix for sure.
@BlenderAddict Жыл бұрын
Hello IS it librairies i have to but or it IS integrated in windows ?
@BobDuffy Жыл бұрын
Its built into Blender. Just download 3.6 or greater and Embree is an option for Arc GPUs using the oneAPI tab
@tseng67 Жыл бұрын
hello thanks for your video, and i have question, when you choose stable diffusion in GIMP, where you find the windows( 5:47)?
@BobDuffy11 ай бұрын
When you select Stable Diffusion from the Layer menu it will open this Window. GIMP is a bit notorious for having Windows pop up in odd places and potentially not on the desktop you have in focus. When recording the video I noticed the window was on another monitor and I dragged it over.
@manikrishnareddy5276 Жыл бұрын
Hi Sir, Did blender 4.0 version support intel xe graphics?
@BobDuffy Жыл бұрын
No. Here's why. Intel Xe integrated graphics was supported in Blender using the OpenCL API in version 2.93. However Blender decided to drop OpenCL starting with Blender 3.0, thus removing its support going forward Intel then added oneAPI as an API in 3.3 which supports Intel Arc GPUs
@manikrishnareddy5276 Жыл бұрын
@@BobDuffy Ok got it. Thanks a lot. It's helpful.
@jettro8523 Жыл бұрын
would love to see a more detailed video of this aspect of the SD gimp plugin
@preshengovender5554 Жыл бұрын
after selecting all of the above I get this error: ModuleNotFoundError: No module named 'diffusers'
@BobDuffy Жыл бұрын
Darm. Be sure to be you are using the recommended version of Python. I believe it says 3.9.13. Meanwhile Ill alert the dev. Always possible something was pushed and creating an issue
@Trick-Framed Жыл бұрын
Nice work.
@TheAstronomyDude Жыл бұрын
Super cool!
@adammalley6140 Жыл бұрын
for anyone who doesn't know, you need a python version between 3.7 and 3.11. 3.11.4 did not work for me
@BobDuffy Жыл бұрын
Good to know
@liquidsnake76 Жыл бұрын
It'd be great to see the full node flow - I got something similar setup to what you have here but it isn't right!
@BlenderAddict Жыл бұрын
Hello pretty intéressant, does it working on amd cpu as well ?
@yelowpunk Жыл бұрын
Once the model loads it shouldn't need to load again between uses. Is the original pop up window getting lost under the main one?
@BobDuffy Жыл бұрын
The current version loads once and you're done and support various 1.5 models
@wenqingye6589 Жыл бұрын
This is awesome! I'm wondering can you show your whole nodes? or do you have a patron page to purchase the project file? Thank you!
@monstercameron Жыл бұрын
what causes the initial latency?
@BobDuffy Жыл бұрын
I believe thats a copy and compile. Which has since been cut down. In the video thats about 20-22 seconds. Current version is about 10-12 seconds
@Trick-Framed Жыл бұрын
Great tutorial.
@isometimestrytodothings2375 Жыл бұрын
yea looks pretty real. looked at my scars and my gf's scars and it looks quite realistic. now you should do 2nd degree burn scars, i have a ton of those
@BobDuffy Жыл бұрын
My sympathies. Fortunate for dragons they tend not to burn, so burns on this sculpt would not make much sense. And, again, my sympathies on your injuries.
@Trick-Framed2 жыл бұрын
Alot of info...im sad I missed it. Hopefully I make the next one if you have one
@mubeenfurqanahmed-mfa68732 жыл бұрын
hi, How did you achieve this? In my case ball kept passing through the racket. can you help me in this regard.
@BobDuffy2 жыл бұрын
That happened to me too. Been so long since I used Unity, but I can tell you why it happens. The ball never hits the paddle becsuse collision is happening between frames. The solution is something called ray casting.
@swdw9732 жыл бұрын
So, 4 months later, still waiting for the ARCS, but here's my actual question. What about alternate render engine support like Renderman, YafaRay, LuxRender, Pov-Ray, Mitsuba, Indigo Renderer, V-Ray, and Octane Render? Lack of this support has pushed AMD GPU's off my list as an option. I will say that I saw a discussion about using an ARC 380 as a secondary card for AV1 encoding being workable with the higher end motherboard chipsets (like my X570), and that caught my attention big time. Discussion was using an Nvidia 3060 or higher card alongside a A380.
@a262best52 жыл бұрын
That Guy Thanks, I will look into it later.
@blenderian.3d2 жыл бұрын
nice denoise with Blender. This is helpful.
@BlenderAddict2 жыл бұрын
Great so please as soon as you have some or you are allowed to publish thèm, Do it. Thanks for advance
@BlenderAddict2 жыл бұрын
Hello so blender is well working on Intel arc but at what kind of speed ? I mean at Thïs Time versus a rtx 3060 or 70 or 80 Inès. I am using blender and i plan to build a New pc maybe with Intel cpu too
@BobDuffy2 жыл бұрын
No performance data as of yet.
@Eradifyerao2 жыл бұрын
I Remember OpenCL - that was when Blender acknowledged my GPU as a rendering device....
@BobDuffy2 жыл бұрын
They still do, but they sorta told AMD, Apple and Intel, "why cant you be like your brother Nvidia and have something like CUDA". So now you'll get an API per manufacturer. Some render engines still support OpenCL, but not sure how much longer.
@twainrocks477111 ай бұрын
@@BobDuffy it's kind of sad how Nvidia is the only option when it comes to serious 3D rendering. Nvidia being the leaders wouldn't be a problem if they didn't charge so high for even entry level GPU's. A 4060 ti isn't worth 410 USD.
@BobDuffy11 ай бұрын
@twainrocks4771 its not the only solution. Serious rendering was done 5 years ago. Now you can do that same rending on low cost hardware and it doesn't need to be Nvidia. You can use a Arc A770 and get performance you had only dreamed of a few years ago, and that gpu is a fraction the cost. Use 2 Arcs if you got $500 to spend on GPUs.