Recently I released my embedding toolkit which has a variety of tools and fixers I've come up with over the past 6 months. I considered adding my emotion embeds to that listing, however I want to keep that specific to helper tools, so I've decided to break this emotion pack into it's own listing.
I've designed these embeds to be lightweight and easy to use. Just add the emotion pack you want right next to your subject name as follows:
a portrait of a Happy512 clown
While they will likely still work if you plop them in other places in your prompt, they work best when they are placed just preceding your subject.
As with my toolkit, your results may vary. I make no guarantees, however I have been using these for awhile now and they seem pretty consistent, especially when mixed with my other embeddings available on my civit page. Follow me here on Civit to be notified when I release new embeddings!
--------
To use these embeddings in AUTOMATIC1111, copy them to your embeddings folder in your stable-diffusion-webui directory. No need to restart, just invoke the embedding from your prompt. Make sure you are using a 1.5 version compatible model or you will get errors!
--------
Description
Holy shit!!!! (this is what I'm trying to capture)
FAQ
Comments (16)
This looks very nice and something I have been looking for to using in my art. - Question... Does this Embedding only contain the "happy512" or does it also contain the other face expressions?
Thank you!
and congratulations on your art!
check version, its there
You'll need to download each version separately, just click through the versions and download each to your embeddings directory. Once there, you can invoke from your prompt just by typing the embedding name.
@socalguitarist Thank you so much good friend ! I will give it a try and see what cool "Happy Faces" it creates :P
What makes a hero? Is it the power they wield? Is it the impression they make on others? Is it their strength in overcoming challenges, or their perseverance in the face of impossible odds.
No.
What makes a hero is how they contribute to those who need them. How they help those limited by the Vram, or their GPU... or even their own HDD space.
This collection... this effort that will benefit so many others, THIS is the work of a hero.
Anyway, I'm off to make pictures of angry Pandas wielding light-sabers!
Great idea. My friend and I came up with a few others really difficult at this time... tired, drowsy, sleeping, weak, unconscious, exhausted and TERROR!!!
Great suggestions, I'll see if I can figure any of those out. If I do, I'll add them here!
Great job, very impressive! But I'm still really curious: How to train a feeling? By using a style rather than a person? How many photos are needed? Only close-ups of interesting facial parts or photos of expressive faces? Possible to have some experimental leads please?
And I think you don't use A1111 for the creation of your embeddings? 🤔Is there a difference between Happy512, Smile512, and Grin512?
There is, though it's subtle. Happy and smile are closest, grin pushes bigger smiles. If you x/y plot them, you'll see a difference, though it's admittedly not a huge difference.
這效果真不錯，感謝分享
These are amazing.
Some notes, when using trigger tokens, the Nervous512 actually registers as Shocked512 and is not too much different than the similar named Shock512. Also, the Grin and Smile should really be swapped since a Smile should really be more intense than a Grin and that intensity is reversed.
This works really well. TYSM! Please make more!
Nice work! Is there any chance to add surprised? Like "Ooopsie"?
these are so good and I haven't found anything else like it. Any chance that they will come to SDXL?
Need XL/Pony ver