GlitchHunter SD15 brings Flux’s color tone and style to Stable Diffusion 1.5, along with GlitchHunter Flux LoRA’s erratic character designs. Trained on 130 images, including GlitchHunter’s training data and additional output data, it blends abstract 2D comic art with 3D sculptural elements and a cyberpunk-glitch aesthetic. The primary goal is to bring the Flux and GlitchHunter aesthetic to AnimateDiff animations. I find it especially useful when working with Dreamshaper for upscaling or generating animations.
Tested with Steerable Motion at 0.5 strength and with latent Upscaling workflow. I use 1.0 strength on LoRA while upscaling and it adds really nice colour and texture. For image generation i got best results with CFG at 4-5 and around 30 steps.
Crediting me while using LoRA is not required but always appreciated. Have fun hunting for glitches and bizarre creatures 🍬 ❤️
Description
Initial version blending 3D, abstract comic, and glitchy toon-style characters. Trained on 130 images from the GlitchHunter LoRA dataset and its Flux outputs.
The goal is to bring Flux’s style to SD1.5, offering better consistency and improved model response when animating or upscaling Flux and MJ outputs.