04.11.2025 - 3.1 Anime style focused version released.
Trained focused on anime-styled arts of Grimphantom with 78 relevant images
Works fine with 1.0 weights
Activation prompt: grimpani
Text extra prompts: boing, smack, jiggle
21.01.2025 - Updated to SDXL Illustrious
Now can generate even more style-accurate images in much better quality.
Retrained with the same data set on the SDXL Illustrious
Activation prompt: grimpcart
Works better with ~1.0-1.1 weights. It is kinda flexible and can be lowered or raised depending on other ised LORAs.
16.09.2024 - New version 2.1 released. This one is focused more on western-cartoonish/comic oriented enad more exaggerated style of Grim.
Activation prompt: grimpcart
Works fine with 1.0 weights
08.06.2024 - Updated to SDXL Pony
Activation prompt: grimphantom
Artist's style works better with more exagerrated style, so try setting higher weights.
1.2 weight for casual generation
Up to 1.5 weights for generating with some heavier character and concept LORAs
Useful body shape prompts: curvy, thick thighs, wide hips, narrow waist
Useful emotion prompts: angry, smug, smile, open mouth
Useful stylistic prompts: digital art, digital artwork, thick lines
Stylistic negative prompts: 3d, shading
Grimphantom artist style LORA. Works better with higher weights, recommended are 1.1-1.2, sepending on checkpoint to reach more accurate style.
Trained on Naifu with 103 images
Main prompt: grimph