Loading
Vous avez un workflow ComfyUI ? Obtenez une documentation lisible directement dans votre navigateur.Rendre un workflow lisible.Essayer maintenant ↗
By now I'm trying to focus on having the most realistic details as posible for general usage, as you can see they see quite different but it's the same model with same keywords, but for the 2 embeddings I'm...
Source du modèle
Description de la source
By now I'm trying to focus on having the most realistic details as posible for general usage, as you can see they see quite different but it's the same model with same keywords, but for the 2 embeddings I'm developing, 1 for normal prompt and the second one for the negative prompt. Plz Test,Review,Post it over here!
Installation : Just download, copy & paste it on your "embeddings" folder, for automatic1111 web ui, is in the root directory.
Usage :
In the normal prompt you can type " CR1 "
In the negative prompt you can type " CRN1 "
I strongly recommend that, if you're looking for human/girls stuff, use both, but if you look for animal, surreal, fantasy or other, use only the negative embedding.
Sources
1 sourceArtefacts du modèle
2 artefactsExtraits de sources
1 extraitSource context: 639 downloads · Type TextualInversion · Base model SD 1.5
Perhaps there will be testing that you should lower scale the embedding or remove it to achieve your goal, will depend in many things, but that's why I'm sharing and want to see your results so I can try to tweak according the images you might post here.
Note : you can try minor scaling since I see that some models doesn't like that yet, please test it as you want
Note of the Note : If you test it, please post it here to see how works on your side!
