<h1 id="saturday-morning-z">Saturday Morning Z</h1>Presenting Saturday Morning Z, a Z-Image Turbo LoRA that captures the energetic charm and clean aesthetic of modern American animation styles.This LoRA is perfect for creating dynamic, expressive characters with a polished, modern feel. It's an ideal tool for generating characters that fit into a variety of projects, from personal illustrations to concept art. Whether you need a hero or a sidekick, this LoRA produces characters that are full of life and ready for fun.Workflow examples are attached to the images in the gallery, just drag and drop the image into ComfyUI. If you see too many artifacts try omitting the trigger and just describe the scene with "cartoon" or "simple illustration" instead. I recommend giving an LLM an example image of a look and having it describe the style, simple prompts seem less interesting.This LoRA was trained in ai-toolkit, stopped at 1,500 steps trained with ~70 AI generated images that were captioned with Joy Caption. Still experimenting with training illustration styles so please understand this is just a early stage version, expect a v2 with better performance.Notes: I've uploaded a 1,250 and the 1,500 step version, I prefer the 1,500 step count, but some might prefer the milder styling of the 1,250 step version.

Saturday Morning Z

Saturday_Morning_Z_Image_Turbo_v1_renderartist_1500.safetensors

v1.0 - Z Image Turbo 1500

v1.0 - Z Image Turbo 1250

Model descriptionPresenting Saturday Morning Z, a Z-Image Turbo LoRA that captures the energetic charm and clean aesthetic of modern American animation styles.This LoRA is perfect for creating dynamic, expressive characters with a polished, modern feel. It's an ideal tool for generating characters that fit into a variety of projects, from personal illustrations to concept art. Whether you need a hero or a sidekick, this LoRA produces characters that are full of life and ready for fun.Workflow examples are attached to the images in the gallery, just drag and drop the image into ComfyUI. If you see too many artifacts try omitting the trigger and just describe the scene with "cartoon" or "simple illustration" instead. I recommend giving an LLM an example image of a look and having it describe the style, simple prompts seem less interesting.This LoRA was trained in ai-toolkit, stopped at 1,500 steps trained with ~70 AI generated images that were captioned with Joy Caption. Still experimenting with training illustration styles so please understand this is just a early stage version, expect a v2 with better performance.Notes: I've uploaded a 1,250 and the 1,500 step version, I prefer the 1,500 step count, but some might prefer the milder styling of the 1,250 step version.Trigger wordsYou should use saturd4ym0rning to trigger the image generation.

Files with SHA256: 34007fbd18523283e3babf5e56e2e08ba292e3ac95ff2325ef28ac179e899a0d