Instructions to use ViraFrame123/ViraFrame-LiNinja-ZImage-LoRA with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use ViraFrame123/ViraFrame-LiNinja-ZImage-LoRA with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Tongyi-MAI/Z-Image-Turbo", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("ViraFrame123/ViraFrame-LiNinja-ZImage-LoRA") prompt = "NinjinhaLi, close-up professional studio portrait of an adult virtual influencer, long straight black hair with a natural center part, warm medium-tan skin, almond-shaped brown eyes, natural eyebrows, full natural lips, white ribbed tank top, neutral gray background, soft diffused studio lighting, realistic skin texture, subtle natural pores, individual hair strands, direct eye contact, 50mm photography, photorealistic, high detail" image = pipe(prompt).images[0] - Inference
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
ViraFrame Li Ninja โ Z-Image LoRA

- Prompt
- NinjinhaLi, close-up professional studio portrait of an adult virtual influencer, long straight black hair with a natural center part, warm medium-tan skin, almond-shaped brown eyes, natural eyebrows, full natural lips, white ribbed tank top, neutral gray background, soft diffused studio lighting, realistic skin texture, subtle natural pores, individual hair strands, direct eye contact, 50mm photography, photorealistic, high detail
- Negative Prompt
- cartoon, anime, illustration, painting, 3d render, CGI, doll, wax figure, plastic skin, porcelain skin, airbrushed skin, beauty filter, fake uniform pores, excessive makeup, oversharpening, blurry face, asymmetrical eyes, crossed eyes, distorted facial features, duplicate person, low resolution, jpeg artifacts, overexposed, underexposed
Model description
Identity LoRA for the ViraFrame AI Studio virtual influencer Li Ninja (@ninjinhali). Trained specifically for Z-Image using AI Toolkit, with 2,000 training steps, rank 32 and BF16 precision. Designed for photorealistic portraits, fashion photography and virtual influencer content.
Trigger word: NinjinhaLi
Trigger words
You should use NinjinhaLi to trigger the image generation.
Download model
Download them in the Files & versions tab.
- Downloads last month
- 13
Model tree for ViraFrame123/ViraFrame-LiNinja-ZImage-LoRA
Base model
Tongyi-MAI/Z-Image-Turbo