AsianVibesMix (SDXL)

AsianVibesMix is a specialized SDXL-based model fine-tuned to achieve high-fidelity, photorealistic imagery with a focus on Asian aesthetics. It excels at rendering detailed skin textures, realistic lighting, and authentic facial features for portraiture and lifestyle photography.

Model Details

• Developed by: [uhuya]
• Model Type: Diffusion-based text-to-image generative model
• Base Model: SDXL v1.0
• Language(s): English (Prompts)
• License: MIT
• Main Focus: Photorealism, Asian portraits, Fashion, and Artistic expression.

Recommended Settings

For optimal results, the following parameters are suggested:

• Sampler: DPM++ 2M Karras or Euler a
• Steps: 30 - 45
• CFG Scale: 6 - 7.5
• Resolution: 1024x1024 (Square), 896x1152 (Portrait), or 1152x896 (Landscape)
• Clip Skip: 1 or 2

Prompting Guide

Positive Prompt:

(photorealistic:1.3), ultra-detailed, 8k, masterpiece, beautiful Asian woman, cinematic lighting, highly detailed skin pores, soft shadows, sharp focus, elegant attire.

Negative Prompt:

(worst quality, low quality:1.4), (deformed, distorted, disfigured:1.3), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, (mutated hands and fingers:1.4), disconnected limbs, mutation, mutated, ugly, disgusting, blurry, amputation, watermark, text.

Limitations & Disclaimer

Users must adhere to the following:

1. Responsible use of generative tools is required.
2. Do not use this model to generate illegal content, non-consensual imagery, or any material that violates local or international laws.
3. Please respect the Hugging Face Community Guidelines.

---

Show your support: If you enjoy this model, please leave a Like ❤️ on the repository!