ComfyUI AI: IP adapter version 2, create artistic images with the Gestalt laws of perception
HTML-код
- Опубликовано: 3 апр 2024
- In this video, I show a workflow for creating cool images using the new IP adapters and the Gestalt Laws of perception in the prompts. This creative test series was a lot of fun. It's really cool what is possible with the new IP Adapter Nodes and the JuggernautXL-Lightning Model!
Развлечения
![Dimension 20: Never Stop Blowing Up Trailer [Dropout Exclusive Series]](http://i.ytimg.com/vi/xYeXpbHnGEM/mqdefault.jpg)
Sauber - echt interessante Konzepte! Thank you for your ambitious and involving work, much appreciated!
Danke, das Coole an Stable Diffusion und ComfyUI ist, dass sich damit die wildesten Konzepte umsetzen lassen, zumindest meistens.
Thanks, the cool thing about Stable Diffusion and ComfyUI is that it can be used to realise the wildest concepts, at least most of the time.
Thank you so much for the tutorial and the workflows. You're so talented. Your images and styles are so creative. Thank you for sharing!
Thanks for the kind words! And I'm glad you like my ideas! :)
Nice. However, the ClipVision seems to be redundant, as Matteo (the guy behind IPAdapter) has mentioned: "if you use the unified loader you never connect the clipvision input (or the insightface). The input is optional and only required if you use the legacy model loaders (eg: IPAdapter model loader). That might be helpful for advanced users who know what they are doing. Otherwise just connect the unified loader and that's all you need."
Thanks for watching. And yes, I know, but I prefer to learn the whole thing in advanced mode. That means I don't use the unified loaders.
Love the results. Thank you for the video. Would be nice if you could share some links to the model and embeds. 😍
I'm glad you like it and thanks for watching and subscribing. You can find and download the IP adapter embeds I created and used in the video on my website: alienate.de. And you can download the Checkpoint model that I used here: civitai.com/models/133005/juggernaut-xl
Have fun trying it out! :)
@@Showdonttell-hq1dk Thank you so much.
My pleasure.
Great video. Could you please provide a workflow?
Thanks for watching, I'm glad you like it! You can download the workflow from my website www.alienate.de. Either just drag the image from the thumbnail into the Comfy user interface or download the json file called: "ipa_two_workflow".
Hey, after several requests, I have added the IP adapter embeds used in the video to my website: alienate.de. If you want, you can download them together with the workflow. Have fun with it! :)
Hello and thanks for the video. I just have 1 thing blocking me:
at (at 11:30) I see "IPAdapter loading Embeds". I have the node but no model.
Where can I find it please?
Hey, thanks for watching and subscribing. You can download the IP adapter embeds that appear in the video from my website: alienate.de. And then all you have to do is copy them into the ComfyUI input folder. Have fun with it! :)
@@Showdonttell-hq1dk just copying doesn`t work
@@CREATERson I have just tried it again:
- Right-click on the IPA Two Workflow Embeds on my website
- Save links as
- directory: ComfyUI Input folder
- Open workflow
- Select embed files in the IPAdapter Load Embeds node
- Make sure that the Load Image (as Mask) node is connected to the IPAdapter Embeds node at "attn_mask"!
- Queue prompt
This is how it worked just now.
I would be pleased if you could let me know if this worked for you.
Thanks for taking a look.
Please share workflow with the images
What do you mean, with the image from the thumbnail, or all the images I used in the video?
I haven't actually uploaded the thumbnail image to my website yet, thanks for pointing that out.