Sign In

faceListiC

797

Updated: Jun 27, 2025

style

Download

1 variant available

SafeTensor

217.88 MB

Verified:

Type
LoRA
Stats

741

797

255

Reviews
Published

May 26, 2025

Base Model

SDXL 1.0

Training
Steps: 548
Epochs: 10
Usage Tips
Clip Skip: 1
Hash
AutoV2
BD3C84BD50
Trigger Words
1girl, faceListic
default creator card background decoration
Uploads - 285

285

Followers - 1919

1.9K

Created on Civitai

2025-05-26-1615_741494760064071_Original.png

Model Type: SDXL | LoRA
Style: Realistic Faces | Chest-Up Portraits
Focus: Beautiful, clean female faces with elegant lighting and natural proportions.


This LoRA is trained on high-quality chest-up portraits of women, focusing on natural facial beauty, cinematic lighting, and photorealistic skin texture. Perfect for creating character portraits, OC renders, fashion-like shots, or stylistic crossovers using semi-realism or painterly styles.

Works best with checkpoints like:
🧠 realvisXL_V4.0, juggernautXL, dreamshaperXL, or any realistic SDXL model.


  • Weight: 0.6 – 0.8 for realism

  • 1.0+ for semi-stylized results

  • Resolution: 1024x1024 or higher

  • Enable Highres Fix if image looks soft

  • Use Euler a or DPM++ 2M Karras for clean structure

  • In my settings for realism I use heunpp2 with SGM_Uniform on both the sampler+scheduler and upscale settings.

  • Heun, Deis, heunpp2, DDIM + ddim_uniform are also viable choices.
    These depending on the checkpoint you are using, I am currently testing this Lora on anime based checkpoints.