Видео

where can I download the PDF??

Key insights 🎨 The combination of image source, sketch creation, and prompt integration in ControlNet creates a new image, making the process more dynamic and creative. 🖥 The local web server will run on your computer, making the installation process more accessible and convenient. 💡 To make ControlNet work locally, you need to download and paste the pre-trained libraries on your computer. 🚀 The installation process for ControlNet and Stable Diffusion is straightforward and can be easily restarted if needed. 🔍 Choosing the right model and prompt can make a significant difference in the results of the segmentation process, highlighting the importance of careful selection. TLDR: The video provides a step-by-step guide for installing ControlNet and Stable Diffusion for creating sketches from images and merging them with prompts, as well as tools for segmentation and depth extraction. 1. 00:00 📹 ControlNet and Stable Diffusion provide a step-by-step installation guide for creating sketches from images and merging them with prompts, as well as tools for segmentation and depth extraction. 1.1 The video provides a step-by-step installation guide for stable diffusion and control net, with a follow-up workshop recording for Penn State architecture department. 1.2 ControlNet enables users to create sketches from images and merge them with prompts to create new images, and also offers tools for segmentation and depth extraction to make the story more interesting. 2. 01:51 📝 Install Python and the Stable Diffusion application by Googling and following the links to download and install the necessary files, then upgrade as needed. 3. 03:09 📦 Install opencv python library and upgrade package manager using CMD in the same folder. 4. 04:36 🔧 Install ControlNet on local stable diffusion by navigating to the folder, running the installation, and accessing the web server, then adding ControlNet from GitHub to the local stable diffusion web UI. 4.1 Navigate to the folder, run the installation, copy and paste the address into your browser to access the local web server for stable diffusion. 4.2 Install ControlNet on top of local stable diffusion by going to the extensions tab, searching for ControlNet on Google, going to GitHub, copying the URL, and installing it from the URL in the local stable diffusion web UI. 5. 06:37 📦 Download the full trained models from the provided link and paste them into the models folder in the extracted folder. 6. 07:37 🔧 Install unstable diffusion, enable control net interface, and restart web UI if issues arise. 6.1 Install the unstable diffusion, apply and restart the web UI, and if there are any issues, shut everything down and restart the web UI and stable diffusion. 6.2 Enable the control net interface in order to use it with stable diffusion, as it is hidden and needs to be expanded in the tab. 7. 09:58 🔧 Installation of ControlNet and Stable Diffusion Local is quick and easy with a powerful Nvidia card, making running segmentation models straightforward. 7.1 Select the preprocessor and model tab, choose the depth model, select the pre-trained model file with the same name, enable control net, and generate. 7.2 Installation of ControlNet and Stable Diffusion Local is fast with a powerful Nvidia card, and running segmentation models is straightforward. 8. 11:45 📹 Install stable diffusion and control net with critical thinking, learn about AI, and support the channel for more content. 8.1 This video provides a step-by-step guide for beginners on how to install stable diffusion and control net, with the suggestion to be a critical thinker while using these tools. 8.2 Learn about AI and how it can change your life, check out the speaker's articles on medium.com, and support the channel for more videos.

Hi, i really enjoyed watching this lesson and i would be really happy if you give us the access to the link resource in order to replicate your exercises. Thank you for your great job

anybody mention minimum GPU requied

This video was absolutely helpful, I’m surprised I’m just coming across your channel now, but better late than never. Thank you for sharing your knowledge and inspiring us

Hello, can you please provide the workshop files for us to download

İçerik için çok teşekkürler.. Bir sorum vardı; bilgisayara yüklediğim controlnet canvas kısmında çizim yapmaya çalıştığımda çizgiler gözükmüyor, bunu nasıl düzeltebilirim?

You're amazing, thank you!

Great conversation

çöp

Great Explanation. Thank you so much !!

'.pip' is not recognized as an internal or external command, operable program or batch file. That's what I get when executing the command

Hi Mr. Gun, i have registered, hope to catch your session on time. have a great day!

What folder are you putting controlnet etc into? You say 'this folder' and it's really hard to follow

Hi Sir, i followed the video to the final step but i cant generate an image. A blurry image appeared when it is generating but when it is completed, this pops out: OutOfMemoryError: CUDA out of memory. Tried to allocate 128.00 MiB (GPU 0; 4.00 GiB total capacity; 2.97 GiB already allocated; 0 bytes free; 3.34 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF I dont know what does it mean, anyone can help me

Many thanks for your great video! Unfortunately, I am the 1%. I received the warming RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check. Would be grateful if you could help me to fix it.

Thanks for the guide! I was using it for my second instalation ) because first was failed =). And do you know it is possible to use stable diffusion and control net for generate animations from videos? this is my goal for now ) and I hope Ill get it soon ))

the extension step was a gem! thanks!!

amazing step by step guide, very helpful, thank you!

when I reopen the webui file it tells me this { ERROR: Error [WinError 2] The system cannot find the file specified while executing command git version ERROR: Cannot find command 'git' - do you have 'git' installed and in your PATH? [notice] A new release of pip available: 22.2.2 -> 23.1.2 [notice] To update, run: D:\AI rendering + rhino\automatic1111\stable-diffusion-webui-master\venv\Scripts\python.exe -m pip install --upgrade pip} .... I wonder if you can help me. Thank you again for your time and efforts

thx for the infos

in minute 9:20 For me, the control is Net 1.1.147 . Is this wrong?

Thanks to you, I was finally able to install and use it👍

hey this is amazing, thanks! could I please have access to the files, would like to try some of the grasshopper scripts for a university course I have. h4m0k4n

At 5:00, I follow the instructions and run the batch file. This is the error message: Exit code: 1

Movie ready materials! It’s impressive looking forward to see what is next :)

Great video content! My only issue is related to the last steps showed. When stable diffusion is started I can only select the preprocessor and not the model in the control net tab. In the video you have multiple solutions for this one (for example the control_sd15_detph), but my only one is "none"". Do you know how to fix it?

Thank you so much, I was struggling with the pip update notice, and finally, it's working now ;) however I had to install the latest stable version of Python, not the latest update to get it work, and while installing it, make sure to enable the "Add Python to PATH" checkbox Otherwise, the batch file may not find the Python installation.

for some reason when I open up the controlnet interface in the Webui I dont see anything under models (just none). The preproccessors are there but not the models.

Great Video, is there a link to you architectural workshop?

“It's about quality, not quantity” The guide is perfect and needed! Keep it up 🔥 always great to see a new video coming out!

Thank you, greetings from Egypt❤️

Does requesting access actually work?

Thank you very much, very very informative workshop✌

I applied for access, no luck yet.

Onur is an amazing teacher. I appreciate his ability to express these explorations so quickly.

Pardon my intrusivess, I did not know where to reach you, here or on Linkedin. So, in short, ACE - architects concil of Europe has a call for researchers in AI/architecture to submit some research directions on the current scene in both domains and I really feel this is right up your alley. Just wanted to bring this to your attention, maybe it is of interest for you. I have been keeping up loosely with your articles on medium or bootcamp and I think your expertise would be a good match. just as a disclaimer: I am not affiliated or endorsed by ACE in any way. Have a great day!

Watching this recording in 2023 gave me two feelings/thoughts by the end of it, - How quickly the tools we are using today have improved; we keep jumping to the next best/new invention/Trend... From Grasshopper to AI makes me wonder what will come next. - Nowadays, we are seeing the effects of AI on the field of Architecture; what you said regarding that in your last part, "Value or Replace", is happening exactly the way you said it, although not many designers have lost their jobs yet to AI, those who got a job as a prompt engineer, now found themselves out of a job thanks to tools like GPT-3, and concept artist/designers added more value to themselves using tools like MJ and Dall-E. New questions came to mind with these rapid innovations of tools like Midjourney and/or GPT-3, some people, including me, are feeling the need to "Rush to learn"; sometimes, I think this feeling comes from the fear of replacement or being left behind and/or excitement. Are those feelings normal, or should we really have those fears? Or are they just a temporary reaction to what's coming? Sometimes we forget how important it is to develop soft skills like critical thinking, Emotional intelligence, Judgement and decision-making. (quoting your post on medium.com, " Complete List of Skills You Need for Enduring Success") When watching the people around me, especially the students and instructors at my school, they mostly seem more worried about how they could improve themselves using AI than how they could improve their awareness and criticality. Just writing my thoughts here. I would like to hear your thoughts regarding what I've said.

As always, you always succeed in surprising me with your critical constructive thinking and work! Keep it up! I love the style of your videos and what you have done with ChatGPT+MJ. P.S.: This could be me only. For some reason, your voice isn't in-sync with your Webcam. I could be just wrong and tired. =]

Hey, great video!! I'm having a little trouble trying to run the google colab. Instead of running it gives me this output: sed: can't read /usr/local/lib/python3.7/dist-packages/gradio/blocks.py: No such file or directory sed: can't read /usr/local/lib/python3.7/dist-packages/gradio/blocks.py: No such file or directory sed: can't read /usr/local/lib/python3.7/dist-packages/gradio/blocks.py: No such file or directory sed: can't read /usr/local/lib/python3.7/dist-packages/gradio/strings.py: No such file or directory Traceback (most recent call last): File "/content/gdrive/MyDrive/sd/stable-diffusion-webui/webui.py", line 7, in <module> from fastapi import FastAPI ModuleNotFoundError: No module named 'fastapi' I don't have any knowledge in coding so can you help please?

The portrait’s fusion with the solo…👌❤️‍🔥

Do they know there are apps like Ms Paint that allow you basic drawing and painting AND it they didn't became inevitably Photoshop?

Yes please!

Brilliant presentation, thank you

yesss, tutorial please 🔥🔥🔥🔥🔥

Sure, a tutorial sounds great

Me, most definitely. Just got in on your ride.

Thanks!!

Not sure if anyone else uses this feature but it really works great: When you've received your downloaded image from MJ, and it's in PNG format, open it up in Photoshop and save it as a jpeg in whichever location on your desktop that suits you best. Next, open up Adobe Bridge. Once Bridge has opened, set the view to "Film strip". Now either drag the JPG file that you have just created into Bridge or locate it via Bridge through the left-hand panel. Once it appears as a thumbnail in your "filmstrip" it will automatically show as a bigger version above in the image pane above, for Mac users: control-click on the image in the image pane (not the thumbnail) - it will bring up a palette box, scroll down to "open up in Camera Raw" which is all the way at the bottom of the palette box. Now what you will see is what looks very similar to what you may have seen in the video, except there is a little hidden easter egg that Camera has for Bridge that it does not have for photoshop, and this is due to the handling of RAW files. All you now have to do is: Control-click on the image and a palette box will pop up, scroll down to where it says: "Enhance", and click - "Enhance", this will now bring up another dialogue box showing a scaled-up cropped section of your image and a "enhance button" click the "Enhance" button. Now if you look at the bottom left thumbnail pannel you will see that there are 2 images, the first image is the original image, and the second image is the enhanced image (x 2 and an algorithm has been applied). You can now tweak the enhanced image with the sliders on the right-hand side accordingly. Click "Open" at the bottom right-hand side of the Bridge application - this automatically opens it in Photoshop. The image will now basically be double the size of the original file and will be in a DNG (RAW) file format. Now here is a little hack that I played around with, the AI in the Bridge will not let you just continue to enhance the image once you have enhanced it once. So if you have enhanced it and opened it up in photoshop, save that open DNG file as a PSD and then save a copy as a jpeg. Now open that jpeg that you have just saved in Adobe Bridge again and follow the process I described above and you can enlarge it again and get a fantastic result because the AI in Adobe Bridge does a really good job. One thing to note is that on your first "Enhancement" do not "sharpen or add any clarity or texture if you do this, the 2nd enhancement will inherit those properties and most likely double them and make your image look over-sharpened or a bit too gritty. On some images I have been able to push out x8 upscale that looks amazing! However the most regular ones have been x4 upscale. It is best to play around, but also to understand your image and know that not one thing fits all images. You can save most of your image settings in Bridge also, so you can start building different templates for the various types of images you enhance. Hope this makes sense and it helps.

Great video, thanks for making. I’m fascinated by this topic, but I’m unsure what the value of the output is. What do you use the generative image for, is it an end product or is it an inspiration for other art forms?