About this LoRA:
Trained on a mixed artist dataset for a consistent, easy to replicate style.

Version notes:
1.0a Base model with initial dataset.
1.0b Tuned for spicier details (Asura-heavy focus).
1.0c Face optimized. Dataset expanded to 300+ images.

Important: These arent linear upgrades. Each variant has unique strengths details that pop in one might be subtler in others.

All versions handle Hires. fix well at high denoise.
For results matching the previews, use 4xNomos8kDAT @ denoise 0.55