My first attempt at creating a styled LORA for Z-Image Turbo. It's based on 80s comics by Spanish artists.
Although it works at strength 1, it can produce more artifacts, so lowering the intensity to 0.9 or 0.8 might be helpful.
For best results, start the prompts with "comic style".
Description
Version 2.0:
This version has been trained on ZImage Base and works perfectly in both ZIB and ZIT.
It is much more consistent and follows prompts more faithfully and with less distortion than the previous version, maintaining the style with greater precision and detail.
The LoRa strength depends on the prompt, but it can be adjusted within a range of 0.7-1 for both base and turbo modes. An acceptable value is 0.9 for most cases.
I hope it helps generate beautiful artistic images.