That one was a bit more of a challenge!
As a big fan of the style of James Jean it took a few tries to get my first style Lora working.
Still not sure where I will be able to take this one. But it already gives very unique results.
I can only recommend to increase the step count a bit to 30 steps or more or make use of a hyper lora with low weight to get Flux to converge faster.
Recommended strength is 0.5 - 0.8. Works best with min 30 steps up to 50 steps
Please feel free to add your images to the gallery!
Description
First release version
FAQ
Comments (2)
Can i use this with a video2video workflow? what model should I use? and do I have to work with flux? It looks so cool! I really want to use it in my project! But I am quite new to all comfy, so sorry if these are obvious questions.
thank you!
I am sadly not familiar with vid2vid. This lora is for Flux only, but I have another James Jean Lora for SDXL as well. Maybe that one is more applicable for what you have in mind.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.