TruckKun Style Lora V3

A style LoRA I made based on the AI-generated images from the pixiv user TruckKun. I like the style he used for his images but couldn't find out what models or loras he used, so I just made one using his artworks. This lora was trained with 801 images generated by him.

Recommended Weight: 0.8

I plan on training another lora eventually to try and create an improved version of this lora with some more images. For now, this lora will have to do.

VERSION 2 (Pony)

My second attempt at making a TruckKun style Lora, using 1000 images for training since that's the limit for CivitAI. In my opinion, this version doesn't work as well. Lines in the image tend to look more "bolded", if that makes any sense, and the shading isn't as close to the original style. Feel free to experiment with this version if you'd like.

Recommended Weight: ~0.6

VERSION 3 (Pony)

The final version I'll be training on the Pony diffusion model. Pretty happy with how this one turned out. Once again, I used 1000 images to train this Lora, but this time I made some small changes to the Lora training settings.

Recommended Weight: ~0.7