saylorrose_char_v1.safetensors

## What this LoRA does

Character-identity LoRA for “Saylor Rose” on SDXL. Trained for face consistency across poses and angles.

Base model: SDXL realistic (1.0)

Best strength: 0.6–0.9 (Fooocus LoRA weight)

## Trained words (activation)

- saylorrose_char

- (optional outfit) saylorrose_outfit

## Recommended prompt (identity only)

Portrait of Saylor Rose, saylorrose_char, ultra realistic, high detail, soft natural light

## Recommended prompt (identity + outfit + scene)

Portrait of Saylor Rose, saylorrose_char, wearing white lace summer dress with brown leather belt, in a Texas sunflower field at golden hour, ultra realistic, high detail

## Negative prompt

low quality, deformed hands, text, watermark, extra arms

## Notes

- Keep same resolution/aspect ratio between shots for best consistency.

- For new poses with same background/outfit: use image reference or inpaint body only.