This is a LoRA model specifically designed for generating celebrity-style images, with character traits and artistic styles deeply integrated during training. It does not require any trigger words in the prompt to activate the target style. Users can simply input basic descriptions (e.g., clothing, background) to produce high-quality celebrity-themed images, ideal for applications like entertainment promotion and character design

meinv001

[v1.0] The first version，Recommended Settings (v1.0):<ul><li>Distilled CFG Scale: 3.5 </li><li>Sampling method and Schedule type: Euler</li><li>Steps: 20++</li><li>Resolution: 896x1152</li><li>Lora Weight: 0.8～1.0. You have to experiment based on your prompts</li></ul>

linyne.safetensors

v1.0

This is a star-style LoRA model requiring no trigger words

Files with SHA256: ce012e8146844e539a3b40fe548e216d06aa072397d638b55e12e69d7d89e96d