Overview

Reina is an SDXL character LoRA trained on Calmform (SDXL merge, Illustrious-dominant).
It’s designed to work even with minimal prompting: reina_v1tok, 1girl should already produce the character, while still allowing easy edits (hair/eyes/hairstyles).

Trigger

reina_v1tok

Recommended settings

• LoRA weight (recommended): 0.65 – 0.85 (sweet spot: 0.75)

• Strong character lock: 0.9 – 1.0

• More flexible / easier edits: 0.55 – 0.7

• CFG: ~3 – 4 (especially on Illustrious-style SDXL checkpoints)

Notes / quirks

• Hairstyle and eye color are intentionally editable.
  If you want the “classic” look, add: twin braids, red eyes.

• If clothing shows unwanted text/logos, add to negative:
  text, logo, watermark, letters, typography

Example prompts

Minimal (works well)

Prompt
reina_v1tok, 1girl, casual outfit, thigh-up shot

Canon / full lock

Prompt
reina_v1tok, 1girl, twin braids, red eyes, casual outfit, thigh-up shot

Edit hair / eyes (controllable)

Prompt
reina_v1tok, 1girl, blonde hair, blue eyes, ponytail

Messy / candid vibe

Prompt
reina_v1tok, 1girl, messy hair, casual photo, natural lighting

Suggested negative

(worst quality, low quality:1.3), blurry, watermark, text, logo, extra fingers, bad hands, deformed

Model details

• Trained on: Calmform (SDXL merge, Illustrious-dominant)

• Type: SDXL LoRA

• Use-case: Character + controllable edits (hair/eyes/hairstyles)