<h1 id="meet-elysara:"><span style="color:rgb(34, 139, 230)">Meet ElySara:</span></h1><p>a stunning brunette character LoRA designed for photorealistic portraits and lifestyle shots. Trained on high-quality imagery, ElySara features rich dark brown hair, expressive eyes, and natural skin texture that blends seamlessly with SDXL and SD1.5 realism models.</p><p>Perfect for creating consistent AI influencer content, fashion photography, or cinematic portraits. Whether you're going for a cozy café vibe, urban street style, or elegant evening wear, ElySara maintains her distinct beautiful brunette features across different outfits and lighting conditions.</p><p></p><p></p><h2 id="recommended-settings:"><span style="color:rgb(34, 139, 230)"><strong>Recommended Settings:</strong></span></h2><ul><li><p><strong>Base Model:</strong> SDXL Realism / Juggernaut / RealVisXL</p></li><li><p><strong>Strength:</strong> 0.7 - 0.9 (Lower for subtle influence, higher for strict likeness)</p></li><li><p><strong>Trigger Word:</strong> <code>elysara</code> (or your specific trigger)</p></li><li><p><strong>Tags:</strong> brunette, dark hair, realistic woman, female portrait, ai influencer</p></li></ul><h2></h2><h2 id="krea-2-raw-and-other-premium-versions-of-this-model:"><span style="color:rgb(34, 139, 230)">Krea 2 Raw &amp; Other Premium Versions of this model:</span></h2><p><a target="_blank" rel="ugc" href="https://www.patreon.com/TensorVizion/posts/highfashion-krea-167277892">https://www.patreon.com/TensorVizion/</a></p><h2 id="generative-ai-workflows-and-resources:"><span style="color:rgb(34, 139, 230)">Generative AI Workflows &amp; Resources:</span></h2><p><a target="_blank" rel="ugc" href="https://www.kofi.com/tensorvizion">https://www.kofi.com/tensorvizion</a></p><p></p>

ElySara

<p><strong><em><u>Trained on Z-Image Use for Maximum Compatibility.</u></em></strong></p>

ElySara_c1-st4000.safetensors

Z-Image

Flux.2-Klein.4B

Flux.2


# 

## Model Introduction

This model is trained using the [Civision/model training](https://modelscope.ai/civision/modelTraining) environment and computing resources provided by ModelScope.

* Model Type: LoRA
* Base Model: [Tongyi-MAI/Z-Image](https://modelscope.ai/models/Tongyi-MAI/Z-Image)
* Training Code: [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)
* Data amount：25
* Total training steps：4000
* License: Apache-2.0

## Inference Code

Install [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio):

```bash
pip install diffsynth
```

Inference:

```python
from diffsynth.pipelines.z_image import ZImagePipeline, ModelConfig
import torch

pipe = ZImagePipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(model_id="Tongyi-MAI/Z-Image", origin_file_pattern="transformer/*.safetensors"),
        ModelConfig(model_id="Tongyi-MAI/Z-Image", origin_file_pattern="text_encoder/*.safetensors"),
        ModelConfig(model_id="Tongyi-MAI/Z-Image", origin_file_pattern="vae/diffusion_pytorch_model.safetensors"),
    ],
    tokenizer_config=ModelConfig(model_id="Tongyi-MAI/Z-Image", origin_file_pattern="tokenizer/"),
)
pipe.load_lora(pipe.dit, ModelConfig(model_id="tensorvizion/ElySara", origin_file_pattern="ElySara_c1-st4000.safetensors"))
prompt = "a cat"
image = pipe(prompt=prompt, num_inference_steps=50, cfg_scale=4)
image.save("image.jpg")
```


Files with SHA256: 2b520faf8b4a891b11199342b0c865fcce66adfa8991fb7f2a22fc79e1756614