触发词：guofeng wuxia, 3D风格的lora，由200张Qwen2509转3D+SD修复而来的数据融炼的lora， 建议使用双采样出图 第1个KSampler建议：步数:30 cfg:5.0 采样器:er_sde 调度器:simple 第2个KSampler(Advanced)建议：步数:30 cfg:5.0 采样器:er_sde 调度器:simple 开始步数:18。 第1个KSampler也可以不接入本lora，先接你的角色lora，第2个KSampler(Advanced)接本lora以确定风格。也可以2个KSampler都接本lora. 第1张图的提示词如下： The prompt for the first image is as follows: masterpiece, best quality, score_7, highres,ultra-detailed, intricate details, year 2025, newest, sensitive,guofeng wuxia,High-end CG illustration with delicate thick painting, visible brushstroke texture, and vibrant Octane rendering. Extremely high definition. The subject is an aloof, stunningly beautiful woman with sharp, defined facial features. Her hair is styled in a loose, voluminous black side braid with stray strands across her face and at the temples, adorned with exquisite hair accessories. Eyelashes are long and delicate. Makeup features blue eyeshadow with fine shimmer and golden glitter accents, complemented by gorgeous, detailed, high-gloss eyes. The skin is extremely pale white, with overall white complexion and facial highlights. She wears elaborate, transparent cloud gauze embroidered with gold thread. The lighting is a cool-toned dark blue. She holds a flowing blue lingsilk ribbon. Setting is a dim lantern-lit throne. The composition is tightly focused on the character's face. —————————————————————————————————————— Trigger word: guofeng wuxia 3D-style lora, obtained by converting 200 Qwen2509 images into 3D + SD format and then refining the data to create lora. We recommend using dual sampling for rendering. Recommended settings for the first KSampler: Iterations: 30, cfg: 5.0, Sampler: er_sde, Scheduler: simple. Recommended settings for the second KSampler (Advanced): Iterations: 30, cfg: 5.0, Sampler: er_sde, Scheduler: simple, Start Iteration: 18. The first KSampler does not need to be connected to this LoRa; connect it to your character’s LoRa first, and connect the second KSampler (Advanced) to this LoRa to determine the style.Alternatively, you can connect both KSamplers to this LoRa.

[Anima]3D风格 武侠_3D style

GFWUXIA_1.0.safetensors

v1.0

触发词：guofeng wuxia, 3D风格的lora，由200张Qwen2509转3D+SD修复而来的数据融炼的lora， Trigger word: guofeng wuxia3D-style lora, obtained by converting 200 Qwen2509 images into 3D + SD format and then refining the data to create lora.

Files with SHA256: 92e6f0a464ceaaeb4db1fdbbae5ec7019c61c2809821b74fc612b8ab3caf86ac