Download
1 variant available
The Anima Model is licensed by CircleStone Labs LLC. Copyright CircleStone Labs LLC. IN NO EVENT SHALL CIRCLESTONE LABS LLC BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.
Built on NVIDIA Cosmos
This lora trained to capture the cute chibi sticker style of Miratsu 169.
The final training set contains about 273 manually curated and corrected tagged samples. Most of the dataset focuses on chibi characters, sticker-like compositions, simple backgrounds, upper-body framing, exaggerated expressions, and cute character gestures.
Recommended trigger word:
miratsu_style
Recommended weight:
0.8 - 1.0
Use around 1.2 for a stronger sticker-like style
Suggested tags:
chibi, simple_background
PS:
This LoRA was trained with underscore-style tags instead of space-separated tags. For example, use simple_background and upper_body, not simple background or upper body
Since this is my first LoRA, it is not perfect yet. If there is some problems occured, please commet me.
这个Lora的训练目标是复现 Miratsu 169 的可爱 Q 版表情包风格。
最终训练集大约包含 273 组人工整理与修正后的标注样本，内容以 chibi、表情包、简单背景、上半身构图、夸张表情和可爱角色动作为主。
推荐触发词：
miratsu_style
推荐权重：
0.8 - 1.0
如果想要更强的表情包风格，可以尝试 1.2 左右
推荐搭配 tag：
chibi, simple_background
PS:
本模型训练时使用下划线格式的 tag，而不是空格格式，例如请使用 simple_backgroundupper_body，而不是 simple backgroundupper body。
这是第一次训练 LoRA，所以还不算完美。如果使用过程中有什么问题，请给我留言。