## 🧔 Middle-aged Muscular Bear Uncle ## 🧔 中年肌肉熊叔Design Concept / 设计理念： Focused on realistic muscular bear uncles, this LoRA specializes in depicting the mature, powerful, and thick charm of Asian middle-aged men. Through targeted fine-tuning, it strengthens explosive muscle lines, dense body hair, and rugged facial features. The goal is to generate muscular bear daddy images that combine raw strength with authentic life-like presence, perfectly matching the reference photo style you provided.Core Features / 核心特征： Muscular bear physique (thick pecs, defined abs, broad shoulders and back, strong arms and thighs), dense chest and body hair, rugged Asian middle-aged facial features (short hair, stubble/mustache, mature wrinkles, steady aura), realistic 40-60 age appearance, emphasizing strength and masculine appeal.Main Strengths / 主要优势： Excellent muscle detail rendering, highly realistic chest hair and skin texture, stable hand and body proportions, natural and commanding expressions. Highly compatible with Z-Image-TURBO, excels at muscular tension and photorealism for Asian bear middle-aged characters.### 🔑 Trigger Word / 触发词 *<code>B1E2</code>**- Compatible Base Model / 兼容底模： Z-Image-TURBO - Strengthens Asian middle-aged muscular bear features + powerful physiqueRecommended Generation Settings / 推荐生成设置： - Sampler: Euler - Steps: 9 - CFG Scale: 1 - LoRA Weight: 0.8-1.0 - Resolution: 1024x10244K Version Note / 4K 版本特别说明： Some of my models are trained on 4K datasets and support higher resolutions (recommended 1328x1328 or above). Details and skin texture will be significantly better. Pair with Hires fix for best results.### 💡 Practical Tips / 实用小贴士 1. Natural Language Prompt Recommendation / 自然语言提示推荐： Best used with natural-language models like Z-Image-TURBO. Use LLMs (ChatGPT, Claude, Gemini, Grok, Qwen) to polish prompts. Simply describe the muscular bear daddy scene you want and let the LLM expand it. 2. Civitai Online Generation / Civitai 在线生图注意： Works decently with Z-Image-TURBO, but local generation on your PC is strongly recommended for full LoRA strength and muscle details. 3. Trigger Word Reminder / 触发词小提醒： The latter part of the model name is usually the correct trigger. 4. Enhance Asian features → add <code>east asian features, black hair, monolids</code> at the end. 5. More muscular/thick → add <code>extremely muscular, huge pecs, prominent chest hair, thick waist, veiny arms</code>. 6. Pose suggestions: <code>standing confidently, flexing muscles, looking at viewer</code>. 7. For better face/muscle details → add <code>highly detailed, ultra realistic, photorealistic, extremely detailed face and muscles</code>.---欢迎使用和提建议，如有帮助，可以选择支持我（BUZZ），谢谢。I organize LoRAs with Collections. If you are interested in more Asian bear middle-aged / Bara Daddy content, you can check my Collection here: 暂无

Middle-aged Muscular Bear Uncle-----B1E2

zimageturbo-b1e2_c1-st3000.safetensors

v1.0(Zimage turbo ZIT)

v1.0(flux.2 klein 9b)


# 

## 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-Turbo](https://modelscope.ai/models/Tongyi-MAI/Z-Image-Turbo)
* Training Code: [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)
* Data amount：20
* 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-Turbo", origin_file_pattern="transformer/*.safetensors"),
        ModelConfig(model_id="Tongyi-MAI/Z-Image-Turbo", origin_file_pattern="text_encoder/*.safetensors"),
        ModelConfig(model_id="Tongyi-MAI/Z-Image-Turbo", origin_file_pattern="vae/diffusion_pytorch_model.safetensors"),
    ],
    tokenizer_config=ModelConfig(model_id="Tongyi-MAI/Z-Image-Turbo", origin_file_pattern="tokenizer/"),
)
pipe.load_lora(pipe.dit, ModelConfig(model_id="jlhe2016/zimageturbo-b1e2", origin_file_pattern="zimageturbo-b1e2_c1-st4000.safetensors"))
prompt = "a cat"
image = pipe(prompt=prompt, num_inference_steps=8, cfg_scale=1)
image.save("image.jpg")
```


Files with SHA256: 58adc0172c152b9b6268dc43ed3c4fae58a443bab93ac1eed57e0c207762a21f