Update v2

Different trigger: angst design

Best sampler scheduler: dpmpp_sde_gpu karras

Help push style by using negatives: photo, blurry, render

What's different?

• Smaller dataset with flip augmentation

• More responsive to prompts

• Stiff in variety of outputs

• Smaller file size for model

Trimmed down the dataset and cleaned up art via vector conversion. Edges are crisper, less colors, more separated. This version is kind of a side grade. It isn't exactly better than the previous or worse as a whole. Parts are better, like this version is more responsive, but it's also harder to push the style. The dataset has less variety so the model is a bit stiff. It can sometimes be impossible to get the style to push properly even with correct triggering in prompt. IMG2IMG is best for pushing the style on subjects that it just can't generate with prompting alone.

The sampler and scheduler can be really tricky with this model. With some scheduler sampler setups it will look like the model isn't doing anything at all or is just making really bad images, blurry mess of things. Different checkpoints will affect this as well. I found that dpm anything, can provide the best results for trying to get the style right.

Art was trained on a class this time, class was design, where as previous versions were made without using a class token. Thought about trying stencil class instead but it isn't really close because there are multiple colors in the image, between 2-4 colors in dataset art.

Something to note that is useful for this model is if you are good at doing color separations for say, plating or vinyl application, the outputs from this model are near perfect for this use. Using something like Inkscape or Illustrator to extract the content for plating separations is great. I am referring to screen printing applications when talking about plating preparation.

Currently this model does what I want, to an extent, but it isn't great. I might revisit this one at a later time to help make it more versatile while maintaining style without nearly as much headache trying to control the style.

~|~

Hello everyone! This LoRA generates images that are illustrative looking and feel on edge, or possibly even "anxious". I think the art to me can sometimes feel broken but still look beautiful. It often makes hard lines, jagged spiky shading, and tends to do a nice job separating the subject from the background.

Triggering the LoRA: Angst in the positive

Putting "painting" in the negative, helps make the image look less smeared and the lines a little more crisp. Sometimes it might want to make photos and putting "photo" or "photograph" in the negative can help guide it back to the style. The v1-2 version hasn't done this to me but previous versions did, so it could still make photos with certain prompts.

Now technically this LoRA can have an effect on the images without any triggering. Calling on the trigger words just helps get it closer to what I was going for to begin with.

I've only used this LoRA style with SDXL in ComfyUI so not sure how it behaves in A1111. I will say though that having only 2-3 steps through the refiner is best. Too many steps in the refiner tends to erase the style from the images.

About the Angst style LoRA

This style came from my love for the KMFDM cover art done by artist Aidan "Brute!" Hughes. Originally I trained a dataset in SD1.5 that was really cool. Though it seemed limited and I found myself wanting something a little different. So with generations made, using outpainting on the originals, and also doing digital painting on the images myself, I went through an iterative process with the LoRA trainings. The dataset for Angst actually doesn't contain any of the original art by Aidan but rather iterative generations from going through using the different versions and making alterations. In conclusion, it still feels a lot like his style, but aside from all the anger it can make a cute looking rat eating a cupcake.