Mika Pikazo Style LoRA V2

简体中文 | English 使用前阅读 | Read before use

--

https://o365cn1-my.sharepoint.com/personal/cdn-cn_o365cn1_onmicrosoft_com/_layouts/52/download.aspx?share=ER_PU4vmpIJOkjEmbTY_eMcBYDawXQFHgs_QRjwta1egSw

--

用于加强画面多彩性的标签：multicolored

Tag used to enhance the colorfulness of the image: multicolored

--

更新：

• 更新数据集，加入新的插画。

• 提升技法拟合效果

• 肢体动作与底模加强拟合，降低面部特写的生成率

• 手部与底模加强拟合，改善对手的拟合

Updates:

• Update the dataset with new illustrations

• Enhance technique fitting effect

• Strengthen fitting of limb movements and base model, reduce the generation rate of close-up facial shots

• Enhance fitting of hands and base model, improve fitting of hands

--

仿Mika Pikazo的作画风格的LoRA模型。

权重建议值：1 (or 0.5~1.2)

基于模型：Anything-5RE

训练信息：共计354张原画(含裁剪图)，迭代约3000步。

拟合比率：

• 肢体拟合：18%

• 人物拟合：84%

• 构图拟合：43%

• 技法拟合：112%

其他的一些建议：EasyNegative, badhandv4。此外，万一畸形四肢换个种子重画就行。

--

The LoRA model mimics Mika Pikazo's painting style.

Recommended weight value: 1 (or 0.5~1.2)

Based on model: Anything-5RE

Training information: A total of 354 paintings (including cropped images) with about 3000 iterations.

Fitting ratio:

• Motion fitting: 18%

• Character fitting: 84%

• Composition fitting: 43%

• Technique fitting: 112%

Other recommendations: EasyNegative, badhandv4. Moreover, in case of distorted limbs, try starting over with a different seed.

--

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.