么懒懒 Style LoRA

中文 | English 使用前阅读 | Read before use

--

最终成图需要较大分辨率。

The final image requires a larger resolution.

--

仿么懒懒的作画风格的LoRA模型。

权重建议值：0.7~1.1

基于模型：Anything-5RE

训练信息：共计294张原画（含裁剪图），最后一次训练迭代1176步。

拟合比率：

• 肢体拟合：17%

• 人物拟合：25%

• 构图拟合：10%

• 技法拟合：110%

其他的一些建议：EasyNegative, badhandv4。

LoRA model with a painting style imitating MeLanLan.

Recommended weight value: 0.7~1.1

Based on the model: Anything-5RE

Training information: A total of 294 paintings (including cropped images) were used for training, with the final training iteration completed at 1176 steps.

Fitting ratio:

• Motion fitting: 17%

• Character fitting: 25%

• Composition fitting: 10%

• Technique fitting: 110%

Other suggestions: EasyNegative, badhandv4.

--

任何问题，欢迎反馈，问题不提出就不会得到解决。如果你创作出了好作品，欢迎分享它。如果它无法满足你的需求，我是说，我的朋友，请不要给出一星好评。

Any questions are welcome for feedback, as problems will not be solved without being raised. If you have created a great work, please feel free to share it. If it cannot meet your needs, my friend, please do not give a one-star rating.

--

题外话

经过试验后我发现较大体积的模型能够轻松提供好的拟合效果，但是每个画风模型都144MB真的能接受嘛……大模型训练起来比小模型轻松不少，哪天我摆烂了可能会直接改用144MB甚至288MB的模型。

Digression

After experimenting, I found that larger models can easily provide good fitting effects, but can we really accept a 144MB model for each drawing style... training larger models is much easier than smaller ones. If I become lazy someday, I may directly use a 144MB or even 288MB model.