nphSi

/

Z-Image-Lora

like 26

Should be ready tomorrow...

I would like to express my sincere thanks to the German Loras. Please keep up the good work and thank you very much!

Can not promise anything since i have still a lot of already made datasets to train on ZI and actually i am doing my first Flux2-klein lora for testing.
Probably can make one or two in the next days...

I would like to express my sincere thanks to the German Loras. Please keep up the good work and thank you very much!

youre welcome 😄

Can not promise anything since i have still a lot of already made datasets to train on ZI and actually i am doing my first Flux2-klein lora for testing.
Probably can make one or two in the next days...

what a hero you are <3

Added Beth Behrs for all three models, will do America F. and Logan B. later but the others are just out of my loot schema (esp Lois Griffin wtf ) 😀

Feel free to support me at ko-fi

Edit:
I was thinking a bit about Lois Griffin because i never made a comic lora. What do you expect here? A simple comic character lora or more CGI / realistic conversion?

Illustrated characters are really hard on ZImage. I've just trained an illustrated penguin i need for a music video on ZImage vs Qwen. Same dataset. ZImage can't even recognize is a penguin. Qwen is flawless instead and i can even use it for photorealistic gens

We just need a bit patience. Illustration fine tunes will come when/if ZI base is out, quiet sure on level with Illustrious/SDXL.

Can not promise anything since i have still a lot of already made datasets to train on ZI and actually i am doing my first Flux2-klein lora for testing.
Probably can make one or two in the next days...

I'm testing it and it's really very good. I can't wait for more.