A combination of Autoprompt and detailer with Metadata.
Version 3 offers the option to use a different resolution.
Version 4;
Minor changes.
You now can choose
Both Daemon Detailer on
Just one Daemon Detailer on
Disable Daemon Detailer at all
V 5: I added Gemini. You can now choose between Florence and Gemini as Model. For Gemini you need a Google account and a Gemini API. I simply prefer Gemini because it is better and more precise. If you use Gemini, bypass the Florence nodes, just to save time.
The Gemini API you can get here: https://aistudio.google.com
Description
Minor changes.
You now can choose
Both Daemon Detailer on
Just one Daemon Detailer on
Disable Daemon Detailer at all
FAQ
Comments (10)
This really looks promising. Unfortunately I can't get it to work, because I do not have the file Presetpro-Neopan.cube. I have been searching up and down, but could not find it. Can you help me please and tell were I can get this? Thank you in advance!
@dominic1336756: Hey great. Thank you very much. I will give it another try now. :)
@BadBabel There are lots of free cube-files on this site. And you don't have to use this one. It only works for black and white. The Midnight .cube is also great for black and white. Or look here: https://civitai.com/models/949550/10k-luts by https://civitai.com/user/pgc
Thank you very much! I had to figure out what this is used for (it brings a special look to the resulting picture). I am still in the beginning with all this. Thank you everybody again for all your help!
I don't understand why the LoraLoaderStackedAdvanced node doesn't recognize my Loras. I don't have any problems with power loader, but in this case, it uses asterisks and detects very few of my loras.
Honestly I do not know :(
I never had problems.
How is your Lora directory called? Lora or Loras?
@hannahthorn loras/flux/
@dominic1336756 should work, but try loras/_flux. I had problems once and after starting the subdirectories with"_" it worked better.