This is an idea that’s been with me for a long time.
This Pantheon (LoRA) blends both Concept and Style.
The Concept is 'Faceless Gods'. Otherworldly deities, both terrifying and divine, born from the depths of our collective unconscious, dreams, and nightmares.
The Style merges dark fantasy with baroque and surrealism.
Heavily inspired by the work of Peter Mohrbacher , H.R. Giger , Igor Kieryluk and Wayne Barlowe.
Description
It was trained with T2I dataset, so trained as Video with 1 frame.
Can be used in combination with the Low model for T2V, but you want to crank up the strength a bit for visible effects.
Produces quite decent results as a I2V, if you add as a starting point an Image produced with just the Low noise LoRA in a T2I workflow.
FAQ
Comments (8)
I like the concept of htis Lora a lot. As a follower of the Orisha, I dig this.
Oh my faceless god.. Its sooo good maaaan
what is a Zit? (not the pimple)
ZImageTurbo
@ArsMachina ty but how do I use that in comfyui? lol is it just a different base model compared to sdxl 1.0? I apologize for my ignorance. :(
@ZUSIMO you can find the separate files on civitai for the comfy workflow (assuming you have the latest cmfyui there's a new template workflow for Z-Image Turbo); or you can use the ZImageTurbo AIO package. You can find it on either HuggingFace or on here. If you have the space you can try both the f16 and f8 versions and pick the one that works best on your machine and gets you best results.
@alucardnoir941 Thank you. ♥
An unfortunate pseudoacronym.