Sign In

MIA_2026_LORA_VARNEZ_1

22

Updated: Jan 31, 2026

character

Verified:

SafeTensor

Type

LoRA

Stats

76

22

113

Reviews

Published

Jan 31, 2026

Base Model

ZImageTurbo

Training

Steps: 540
Epochs: 10

Usage Tips

Clip Skip: 1

Hash

AutoV2
E957D40B9F

Created on Civitai

This LoRA is designed to generate photorealistic portraits of an adult woman with consistent features: long straight platinum hair with darker ombre ends, fair skin with natural texture, green-gray eyes, and soft makeup (glowy finish, subtle eyeliner, defined lashes). It performs best for close-ups and medium shots in studio-like setups, clean backgrounds (white/gray), soft lighting, and an editorial/beauty aesthetic. It also supports moderate variations in pose, camera angle (front, 3/4, profile), and outfit changes while preserving facial identity and hair style.

This resource is not intended for complex action scenes, crowds, multi-subject compositions, or highly busy backgrounds, where identity consistency may degrade. It’s also not recommended to push it into extreme styles (cartoon, anime, painting, heavily stylized renders) if you want to keep a photographic look. For best results, use it in portrait-focused contexts and keep the LoRA weight moderate to avoid overbaking (exaggerated features, overly plastic skin, or hair artifacts). It should not be used to impersonate real people or to create deceptive content.