Unlocking the Secrets of Automatic 1111 with SD XL Checkpoints
HTML-код
- Опубликовано: 30 май 2024
- In today's video, I take you on a deep dive into the world of Dreamshaper XL, the one of the newest and best SD XL Checkpoints. We'll skip the basic introduction and instead, focus on how to maximize its capabilities, bringing your imagination to life. If you've been curious about how SD XL Checkpoints work, or if Dreamshaper is the right fit for you, this video is a must-watch.
🔥 Key Highlights:
Exploring Dreamshaper XL Checkpoints: Dive into the magic behind extended models and the aim of capturing unique art styles.
Setting Up: A step-by-step guide to getting started with Dreamshaper XL and refining with Dreamshaper v8.
Dragon Rendering Challenge: I put Dreamshaper's capabilities to the test with an award-winning photo of a dragon, testing various samplers and examining their outputs. Will it hold up to its promise of 75% better dragons?
Refinement and Latent Upscale: Enhance the details in your creations using latent upscale and learn the tricks of the trade.
Tile Upscaling: Boost your image's resolution without compromising on the details.
Bonus Content: Discover a treasure trove of prompts and art styles, alongside tips on effectively using them.
Remember to hit the subscribe button if you find this video helpful. Your support goes a long way in helping me produce quality content for you! Drop your thoughts, questions, and suggestions in the comments below, and as always, happy creating! 🎨🖌️
Book a one-on-one appointment:
aiknowledge2go.simplybook.it/v2/
Patreon Free Stuff: / aiknowledgecentral
Ultimate Beginners Guide: / sneak-peek-alert-90799508
Free workflow guide: / get-your-free-99183367
Style Collection: / my-collection-of-87325880
My ComfyUI Tuts:
• ComfyUI
My Automatic 1111 Tuts:
• Automatic 1111 - Stabl...
AI Tools:
• AI-Tools
Resources:
Aspect Ratio Helper: github.com/thomasasfk/sd-webu...
Dreamshaper v7 : civitai.com/models/4384
Dreamshaper XL civitai.com/models/112902/dre...
Styles: / my-collection-of-87325880
Artists: supagruen.github.io/StableDif...
huggingface.co/stabilityai/st...
Chapters
00:00 Intro
01:05 Updating A1111
02:10 Setting the Scene and rendering
05:07 Refinement
06:45 Increasing quality even further
External Resources:
www.vecteezy.com/video/207500..."4K video - Setting temperature on air fryer Stock Videos by Vecteezy
Tags:
#AINews #KINews #AI #KI #AiTools #LLM #chatgpt #gpt #künstlicheintelligenz #artificialintelligence #KIWissen2Go #AIKnowledge2Go #StableDiffusion #ImageGeneration #Bildgenerierung #Automatic1111 #Tutorial #Learn #Rendering #RevAnimated #Civitai #AIArt #KIKunst #HowToGenerateAIArt #Workflow #Image2Image #Inpainting #Upscaling #Upscaler #Inpaint #AIartTutorial #LearnAI #Guide #Howto #InstallSDXL #Automatic1111Tutorial #AI_ArtCreationGuide #SDXL #Dreamshaper, #Mastering, #Master class 
Наука
Do you struggle with prompting? 🌟 Download a sneak peak of my prompt guide 🌟 No Membership needed: ⬇ Head over to my Patreon to grab your free copy now! ⬇ www.patreon.com/posts/sneak-peek-alert-90799508?Link&
Always look forward to your tutorials. Definitely some of the best teaching about Image gen workflows on the Internet. Thanks for making this!
Thank you so much for the kind words! I'm thrilled to know that you find my tutorials helpful. It's feedback like yours that keeps me motivated. Stay tuned for more content, and happy image generating!
Fantastic content. Thank you for sharing!
Thank you so much for the appreciation! Always happy to share and help out. Stay tuned for more! 😊
Thanks for great tips, very useful!
You're welcome! I'm glad you found them helpful. Stay tuned for more! 👍
Warning: May take forever! Nice video
why did you have to load the dragon to 'PNG info' then send it to 'img2img', when you could have sent the image from 'txt2img' in one click? Just curious
I rendered an XYZ plot. If I had sent it over, I would have sent the whole grid. I could have dragged the image into img2img. Loading it via PNG info is just my preferred way of loading images. Hope that helps.
ADetailer (section) or option I don't have on my ui, please help 😢 even if I generate the image it comes out low resolution I can't go higher than 512x512 but I have seen many people use it and get better results, someone suggested me to use ADetailer but I can't and don't know how to
Sorry to hear you're having issues. It seems like you might be having memory limitations. Adding the --lowvram or --medvram parameters when initiating automatic 1111 can sometimes help with these problems. To do this, head over to your stable diffusion folder and open the webui-user.bat. Search for the line "set COMMANDLINE_ARGS=" and add one of those parameters. Remember, after this adjustment, always launch Automatic 1111 using the webUi-user.bat.
As for ADetailer, I've made a dedicated tutorial on how to install and utilize it: ruclips.net/video/y3DxX9s0NhQ/видео.html. I hope this helps!
@@AIKnowledge2Go yep show and got it fixed
Hey, there is not link in the description for the styles csv 😢 it is possible to have it please? :)
Hmm, that's odd. It actually should be there. I added it to the description, but here you go: www.patreon.com/posts/my-collection-of-87325880
@@AIKnowledge2Go Thank you!
I have RTX 2060 and when I use SD 1.5 imges get generated fast at ~1.5s/it but when I use SDXL models it takes long and 10s/it please help❤
the SDXL models are indeed more demanding on VRAM & GPU. Have you tried adding the 'no-half-vae' modifier to your webui-settings? This could potentially improve your performance.
Wait.. why deactivate VAE? Whats it used for?
The VAE is basically an Image Enhancer. When SDXL came out it needed a refiner. It doesn't anymore. So you can just keep using your SDXL VAE. Just don't mix SDXL and SD 1.5 VAEs or you will get strange colors.
do you use InvokeAI?
I primarily rely on Automatic 1111 and ComfyUI. For this particular video, I utilized Automatic 1111.
damn, i don't have the "styles" file you mentioned. i' still haven't watched your other video tutorial, but tbh i think you are going a tad too fast, might you be more explicable in the passages you do you would be more comprehensible.
tho, i have to be sincere, you go really in depth, good work
I appreciate the feedback! Striking the right balance between being too detailed and too brief can be challenging. In the past, some viewers preferred that I don't explain every single mouse click, but I understand where you're coming from. I'll try to find a balance in future tutorials.
Regarding the "styles" file issue: Check in your main "Stable-Diffusion" directory, where the webui.bat file is located. If you don't find the "styles" file there, you can create a new .csv file named "styles". Open that file and paste the necessary content. I hope this resolves your issue.
@@AIKnowledge2Gooh thanks I'm going to try this! I had the same issue. ❤
Where is this reddit leake style prompt post?
Found it r/StableDiffusion/comments/15afvnb/sdxl_various_styles_keywords/
There is a possibility that the post on reddit got deleted, but i got you. You can find the pompts in the link in my description to my patreon. 100% for Free!
If you want only XL prompts just delete the section that says:
____SD1.5 Positve Prompts Only____,,,,
and
____*Negative Prompts Only____,,,,
I literally laughed out loud reading that
Mach es uns beiden einfacher und rede deutsch XD
Klar, hätte ich gerne gemacht! Aber mein deutscher Kanal, KIwissen2Go, hat in der Zeit, in der ich hier 1000 Abonnenten erreicht habe, nur 83 geschafft. Das sagt wohl alles. 😅
@@AIKnowledge2Go War nicht so ernst gemeint. Deine Videos sind wirklich gut und Hilfreich. Du sprichst auch sehr deutlich, so das man gut mitkommt. Ich finde nur deinen deutschen harten akzent etwas witzig, weil ich mich wahrscheinlich genau so anhöre Aber einen echten kleinen Kritikpunkt hätte, ein etwas besseres Mikro würde dir echt gut tun sobald es der Kanal finanziell zulässt. Ansonsten mach weiter so, du hast deutlich mehr Abos verdient.
@@Pabstie Kein Problem, ich habe es nicht als ernstes Kommentar aufgefasst. Ein besseres Mikrofon steht definitiv auf meiner Wunschliste und wird meine nächste Investition sein. Vielen Dank für das konstruktive Feedback und die netten Worte!