MJ Style Distilled 206.safetensors

316 MB · SafeTensor · bf16

Memerlukan keahlian

<mj-style-flux2>

EN： This LoRA was distilled from an MJ-style model, with the goal of reproducing a wide variety of artistic styles in a more efficient and consistent way. The current version includes 206 styles in total. To make it easier to use, I also created a style browser and comparison webpage where you can check the style names, sample images, and side-by-side comparisons between the original model and the LoRA model using the same style prompts: Features - Distilled from an MJ-style model - Includes and organizes 206 styles - Helps you quickly apply different visual styles in your workflow - Comes with a reference webpage for style lookup and image comparison How To Use If you are not sure how to use it, or if you want to quickly find a style that fits your needs, please visit the webpage above. I compared the original model and the LoRA model under the same style prompts, so you can more easily evaluate how each style performs and decide which one works best for your generation workflow. Suggestions - Start with your usual base prompt, then add the style prompt for testing - Use the comparison webpage to quickly browse and filter styles - Results may vary depending on the base model, so adjusting LoRA weight is recommended Support If you find this model helpful, feel free to support me: 中文： 这是一个通过蒸馏 MJ 风格模型得到的 LoRA，核心目标是更高效、更稳定地还原多样化的风格表现。 当前版本共整理并支持 206 种风格。为了方便使用，我专门制作了一个风格检索与对比网页，你可以在这里查看每种风格的名称、示例图，以及原模型和 LoRA 模型在相同风格提示词下的效果对比： 模型特点 - 基于 MJ 风格模型蒸馏得到 - 收录并整理了 206 种风格 - 便于快速调用不同画风进行创作 - 可结合网页查看详细风格说明与图像对比 使用说明 如果你不确定该如何使用，或者想快速找到适合自己的风格，可以直接访问上面的网页查看详细风格和图片对比。 我对比了**原模型**和**LoRA 模型**在相同风格提示词下的表现，方便你更直观地判断不同风格的适配效果。 建议 - 建议从你熟悉的基础提示词开始，再叠加风格词进行测试 - 如果想快速筛选风格，优先查看网页中的对比结果 - 不同底模下效果可能略有差异，建议自行微调权重 支持我 如果这个模型对你有帮助，欢迎支持我：

SHA256

52989688E1A983DF27540F72A7C135C101E587EE46BE9D089C92B5C402F65255

ID Model ModelTrends.ai

#26183

Sampler

Euler

Seed

561714946974088

Saiz

720x1280