Trigger: katewmn

Strength: 0.85–0.92 (start at 0.9)

Original AI-generated character (not a real person).

Kate is an athletic adult blonde woman in her late 30s — gym / lifestyle / glamour photoreal look. SDXL LoRA for consistent identity across outfits and poses.

Trained on an original AI character dataset. Fictional OC only — no real-person likeness.

How to use (ComfyUI / A1111):

• Photoreal SDXL base: RealVisXL / CyberRealisticXL / Juggernaut

• Load LoRA katewmn_sdxl_lora.safetensors @ ~0.9

• Put katewmn at the start of the prompt

• Optional: face restore for hard multipose if identity drifts

Positive template:

katewmn, photorealistic photo of katewmn, athletic woman late 30s, long wavy blonde hair with darker roots, freckled fair skin, [POSE / OUTFIT / SCENE], natural skin texture, visible pores, raw photo

Negative template:

cartoon, anime, 3d, cgi, doll, plastic skin, oversmooth, young adult 20s, deformed, bad anatomy, blurry, watermark, text

Training:

1400 steps · dim 16 / alpha 8 · RealVisXL V5 base · 768 bucketed · AdamW · min-SNR 5 · noise offset 0.035