You can click on the link below to try it out directly. If the effect is good, you can deploy it locally<a target="_blank" rel="ugc" href="https://www.runninghub.ai/post/2086745289859379202?inviteCode=rh-v1559">https://www.runninghub.ai/post/2086745289859379202?inviteCode=rh-v1559</a>Experience this LoRA for free at the link below! Powered by a Krea2 workflow, it enables seamless direct 4K high-definition generation.Welcome to Krea2_ShadowWuxia_Realism!A specialized cinematic LoRA engineered for the Krea2 Architecture, focusing on realistic BJD (Ball-Jointed Doll) male martial artists, dark Wuxia aesthetic portraiture, shadowy cinematic lighting, and traditional fantasy warrior designs.This model excels at combining delicate BJD doll porcelain facial traits with rugged, sharp Wuxia character aesthetics. It captures intricate traditional robes, leather and metallic armor accents, dramatic shadows, and high-intensity atmospheric moodiness. Perfectly suited for generating male Wuxia character portraits, dark fantasy martial arts concepts, oriental game cinematic stills, and BJD doll-style fine art illustrations.Recommended ParametersBase Architecture: Krea2 Running Version: Krea2 Turbo Sampling Steps: 8 CFG Scale: 1 LoRA Weight: 1.0 (Recommended 0.8 to 1.0) Core Trigger Tags: shadow wuxia realism, male wuxia warrior, bjd male doll, ball-jointed doll face, dark wuxia style, oriental fantasy character, dramatic shadow lighting, cinematic portrait, masterpiece renderPrompting &amp; Style Tips<ol><li>Wuxia &amp; Outfit Enhancements: For Martial Arts Attire: wearing layered dark traditional robes, leather chest guard, metallic shoulder armor, intricate fabric textures, long dark ponytail hairstyle For Cinematic Atmosphere: low-key lighting, moody shadows, atmospheric fog, rim light highlights, dark background, sharp sword accent</li><li>Character &amp; Doll Feature Prompts: For BJD &amp; Male Portraiture: ball-jointed doll face, porcelain skin, sharp defined jawline, intense focused gaze, stoic expression, fine facial details, 8k resolution</li></ol>

Krea2_ShadowWuxia_Realism

Krea2_ShadowWuxia_Realism.safetensors

v1.0


# Krea2_BJD_男武侠

## 模型介绍

本模型依托魔搭社区（ModelScope）AIGC专区[模型训练](https://modelscope.cn/aigc/modelTraining)环境与算力完成训练。

* 模型类型：LoRA
* 基础模型：[krea/Krea-2-Turbo](https://modelscope.cn/models/krea/Krea-2-Turbo)
* 训练代码：[DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)
* 训练数据量：63
* 总训练步数：10000
* 开源协议：Apache-2.0

## 推理代码

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

```bash
pip install diffsynth
```

开始推理：

```python
from diffsynth.pipelines.krea2 import Krea2Pipeline, ModelConfig
import torch

pipe = Krea2Pipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(model_id="krea/Krea-2-Turbo", origin_file_pattern="turbo.safetensors"),
        ModelConfig(model_id="Qwen/Qwen3-VL-4B-Instruct", origin_file_pattern="*.safetensors"),
        ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="vae/diffusion_pytorch_model.safetensors"),
    ],
    tokenizer_config=ModelConfig(model_id="Qwen/Qwen3-VL-4B-Instruct", origin_file_pattern=""),
)
pipe.load_lora(pipe.dit, ModelConfig(model_id="zhouwenbin1994/Krea2_ShadowWuxia_Realism", origin_file_pattern="Krea2_ShadowWuxia_Realism_c1-st10000.safetensors"))
prompt = "a cat"
image = pipe(
    prompt, seed=0,
    num_inference_steps=8, cfg_scale=1, mu=1.15,
)
image.save("image.jpg")
```



Files with SHA256: aa40645f5d919786d110d71faa03c73a903f1ca378cd331f986c699462524738