Joana — Z-Image Turbo LoRA

An identity/character LoRA for Tongyi-MAI/Z-Image-Turbo.

Suggested trigger word

Start prompts with joana. This word is inferred from the repository and weight filename; the original training captions are not included, so the exact trigger token could not be independently verified.

Example prompt:

Photorealistic portrait of joana, natural window light, realistic skin texture, detailed eyes, 85mm photography, shallow depth of field.

Recommended settings

• Base model: Tongyi-MAI/Z-Image-Turbo
• Start with LoRA strength 0.8–1.0.
• Use 8 inference steps.
• State the desired pose, clothing, framing, lighting, and setting explicitly.
• Lower the strength if prompt flexibility decreases; raise it gradually if identity is weak.

fal.ai example

{
  "prompt": "Photorealistic portrait of joana, natural window light, realistic skin texture, 85mm photography",
  "image_size": "portrait_4_3",
  "num_inference_steps": 8,
  "loras": [
    {
      "path": "https://huggingface.co/hodgy/joana/resolve/main/joana_z-image-turbo_lora.safetensors",
      "scale": 0.9
    }
  ]
}

File

• joana_z-image-turbo_lora.safetensors — Z-Image Turbo LoRA weights in Safetensors format

The file contains 480 LoRA tensors. Its embedded metadata identifies PyTorch format but does not include the training step count or captions.

Intended use and limitations

Use for character-consistent portrait and creative image generation. Results can vary with pose, distance, occlusion, stylization, and LoRA strength. Hands, text, small facial details, and crowded scenes may contain artifacts. Outputs inherit limitations and biases from the base model and training data.

Responsible use

Use only where you have the necessary rights and permission. Do not use the adapter to impersonate, deceive, defame, harass, exploit, or violate privacy or publicity rights. Clearly disclose synthetic imagery when context could cause viewers to mistake it for a real event or endorsement.