Trained on ~200 images
1024x1024 works best but i've had good results with various resolutions.
The model labeled with 'Less' seems to nail down the style better. I liked how the 'Lots' model showed a bit more variety so i thought to keep both.
Recommended promp:
Positive: cromachina, half-closed eyes, narrowed eyes, eyelashes,
Negative: censored
Description
Details
Downloads
434
Platform
CivitAI
Platform Status
Deleted
Created
6/15/2024
Updated
8/9/2025
Deleted
8/7/2025
Trigger Words:
Cromachina
Files
Cromachina_test-000009.safetensors
Mirrors
CivitAI (1 mirrors)