This style LORA is intended to be paired with a model mix including a CKPT of the same name, you can find that here: djzAgentOrange
This is a specific method of merging base CKPT models and running a stack of these Style LORA on top. You might use this alone, however it was designed to work in symphony with a group of other djz resources
Demo images use Danger Zoo ~ Johnsons Negative TI Collection ~
Description
FAQ
Comments (3)
I'm relatively new and haven't yet come across an example of your intended workflow that I felt like I understood. Any recommended resources to help me get my head around a practical example of how to do this?
Everything I've seen so far looks really cool. Thanks.
i will provide a video guide soon demonstrating how to use the djz resources, the basics are this: take one or more djz models and merge them into each other or a larger model. Then balance the trigger tokens as you would any word in a prompt by using weighting. Then take the style LORA's and stack them on top, again paying attention to the network strength. Combine all of this with a good prompt and solid negative embeddings in 2.1 will give you a powerful concept engine that is "anchored" to the themes contained inside each of the models you chose. Its possible to mis these assets in abstract ways and create original works by your design.
@driftjohnson Thanks. I think I can get my head around it, but I'll definitely look forward to a video guide as well.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.