为Anima base-v1.0制作的实验性通用型编辑Lora，使用小型验证性全动漫编辑数据集训练。
你会需要ComfyUI-Cosmos-Reference来运行该lora
对于难以起效的改变,可以尝试提高提示词的权重。（anima对高权重的耐受性极强，猛猛冲到顶也没关系）
这只是一个初期实验lora,编辑能力低下且难以控制,将来会尝试更多功能不同、更强响应的编辑Lora。
An experimental general-purpose editing LoRA made for Anima base-v1.0, trained on a small proof-of-concept all-anime editing dataset.
You will need ComfyUI-Cosmos-Reference to run this LoRA.
For edits that are hard to achieve, try increasing the prompt weight.
Description
超早期训练lora，你可能遇到任何意料之外的情况
FAQ
Comments (11)
?!强!?
Thanks a lot for this lora! Works great.
I just compared this to my IC-like method.
Looks good! Any idea how this compares to controlnet-lllite?
how to disable original image and straight to output?
this is so fucking usefull, why I didn't find it earlier🤩🤩🤩
where can i find a workflows?
Is there a guide somewhere on how to use this properly?
老大不会用喵，有没有工作流可以分享一下的，想拿来做差分图
if you get garbage: go to ComfyUI\custom_nodes\ComfyUI-Cosmos-Reference and open command line and run git checkout 6db3b2c
the owner messed up his newest version.
How do I do it with ComfyUI portable?
有示例工作流或者教程什么的吗老哥