🔐 IDENTITY LoRA (SDXL) — Clean Face Lock - v1

🔐 IDENTITY LoRA (SDXL) — Clean Face Lock

LORA
Original


Updated:

Version Detail

Trained by Tensor
SDXL 1.0
10
🔐 IDENTITY LoRA (SDXL) — Clean Face Lock by FULIX STUDIO This is a pure IDENTITY LoRA trained to accurately reproduce a single adult female face from real photos, without baking in any style, lighting, or mood. It is designed to keep facial structure consistent across: • Close-ups • Mid-shots • Full-body shots • Any pose or camera distance Style is intentionally NOT included. This allows you to apply any look later (beauty, doll, cinematic, anime-leaning, SFW/NSFW) using your base model or a separate style LoRA. ✨ WHAT THIS LoRA DOES • Strong facial identity & proportions • Stable face across poses and distances • Clean eye structure • Flexible with all SDXL styles • Beginner-friendly prompting 🚫 WHAT THIS LoRA DOES NOT DO • No baked style • No lighting bias • No doll/anime bias • No outfit or pose bias 🧠 BASE MODEL (GENERATION) • Any SDXL base • Recommended: Pony Realism v2.2 for best identity stability 🔑 TRIGGER WORD • <gf_id> ⚙️ RECOMMENDED SETTINGS • Resolution: 1024×1024 • Sampler: DPM++ 2M Karras • Steps: 34–38 • CFG: 5.5–6.2 🎚️ LoRA STRENGTH • Close-up: 0.9–1.0 • Mid-shot: 0.8–0.9 • Full-body: 0.7–0.8 This LoRA is ideal as a foundation. Add style at generation time using prompts or a separate style LoRA. Created by FULIX STUDIO — focused on clean, flexible, high-quality character foundations.

Project Permissions

    Use Permissions

  • Use in TENSOR Online

  • As a online training base model on TENSOR

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated contents

  • Use on generation services

  • Sell this model or merges

Related Posts