这是一个使anima模型增加简单编辑能力的lora，由于训练集主要是剥离衣物的差分，所以主要的功能是剥离衣物，其他编辑能力几乎没有。已测试的额外功能是表情替换、去除审核（训练后期增加了手工合成的相关数据集，但是未充分训练）。图片的其他生成结果请参考我的pixiv主页https://www.pixiv.net/users/99228777

注意：工作流已嵌入到图片中，使用该模型必须配合ComfyUI-Cosmos-Reference节点使用

训练工具：https://github.com/Mirumo0u0/sd-scripts/blob/main/docs/anima_conditioning_training-zh.md

打标教程：Identify differences in the tags of variant images | 获取变体图片的标签差异 | Civitai

This is a LORA model designed to add basic editing capabilities to the Anima model. Since the training dataset primarily consists of variations involving the removal of clothing, its main function is to remove clothing; it has virtually no other editing capabilities. Additional features that have been tested include facial expression replacement and censorship removal (a dataset of manually synthesized images was added during the later stages of training, but the model was not fully trained on it). For other generated images, please visit my Pixiv page: https://www.pixiv.net/users/99228777

Note: The workflow has been embedded in the image. This model must be used in conjunction with the ComfyUI-Cosmos-Reference node.

Training tools: https://github.com/Mirumo0u0/sd-scripts/blob/main/docs/anima_conditioning_training.md

Labeling Tutorial: Identify differences in the tags of variant images | 获取变体图片的标签差异 | Civitai