<span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, &quot;system-ui&quot;, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">This model is intended to imitate Greg Capullo's artstyle. Greg Capullo is the penciller for the New 52 Batman series starting with Court of Owls<span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, &quot;system-ui&quot;, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">. This is a LoRa trained on illustrations by Greg Capullo in a mix of color, ink, and pencil. It's somewhat finicky. Sometimes it really get's the style right, other times it's a little looser. I'll update this description if I find any techniques for getting a more specific style.This is a Z-Image version of my Flux LoRa.

Greg Capullo Style

This is a second attempt at this LoRa using AIToolKit rather than CivitAI trainer. I think it worked a little better.I've been use Euler sampler and Simple scheduler at about 0.6 strength for my generations, which seems to preserve style without destroying anatomy too badly. I have found adding some descriptions of the style can help improve the look of images. For example, "The sketch uses outlines and hatching to indicate shadows and light source. It is a pencil on paper sketch with cel shading. A stylized sketch" or "comic book inspired style" or "The illustration uses outlines and black shadows and light source. It has been digitally inked and colored. A stylized illustration"

Files with SHA256: d6ccef65f3424a367ee40da2a51d141d7af567aef756317b0f81f1c402976145