Mila is a photorealistic character LoRA created specifically for Z-Image Turbo.

She is a fictional adult woman with long straight black hair, a clean center part, pale blue-gray eyes, dark sculpted eyebrows, full lips, a warm light-olive complexion and naturally curvy proportions. The LoRA was trained for consistent identity across close-up portraits, selfies, side profiles, three-quarter angles, varied facial expressions and full-body compositions.

### Trigger

MilaVoss

Place the trigger near the beginning of your prompt:

MilaVoss, adult woman, ...

### Recommended strength

Start at 0.80.

- 0.65–0.75: More prompt freedom

- 0.80–0.90: Best identity balance

- 0.95–1.00: Stronger identity, but more training-set influence

Load MilaVoss only once. Stacking the LoRA twice or combining it with another strong character LoRA can reduce prompt adherence and distort facial features.

### Suggested Z-Image Turbo settings

- Steps: 8–12

- CFG: 1.0

- Sampler: Euler

- Scheduler: Beta

- Z-Image shift: 1.8

- Multiplier: 1.0

- Recommended portrait resolutions: 832×1248 or 1024×1536

A blank or lightweight negative prompt generally works best.

### Strengths

- Photorealistic portraits and selfies

- Stable facial structure

- Frontal, profile and three-quarter views

- Natural and glamour photography

- Different expressions and camera angles

- Close-ups and full-body compositions

- SFW and adult content

### Current v1 limitations

The training set contains recurring black strapless clothing, dark fur styling and simple white/gray backgrounds. At high strengths, these elements may appear without being requested. Eye color and skin tone can drift slightly under unusual lighting. Reinforce pale blue-gray eyes and light warm-olive skin when those details are important.

This is the first public version. Feedback and example images are welcome.