Instructions to use SAkizuki/Anima-Aniart with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use SAkizuki/Anima-Aniart with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("circlestone-labs/Anima", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("SAkizuki/Anima-Aniart") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
Anima-Aniart - 手绘水彩风格 LoKR
一款为Anima模型设计的手绘水彩风格LoKR,旨在复现"Ani-Art"独特的手绘、水彩和颜色漂移美学效果。
关于本LoKR
这是我发布的第一个LoKR模型,专门用于捕捉Ani-Art作品中常见的手绘水彩风格。本LoKR能够生成具有以下特点的图像:
- 手绘美学:有机、不完美的线条,呈现传统手绘质感
- 水彩效果:柔和的色彩混合、自然的颜色渗透和半透明渲染
- 颜色漂移:微妙的色彩偏移,增加深度和艺术感
- 绘画质感:可见的笔触和纸张般的纹理
注意:作为我的第一个LoKR发布版本,可能存在一些不完善之处。我会根据社区反馈持续学习和改进未来的版本。
模型详情
| 属性 | 值 |
|---|---|
| 模型名称 | Anima-Aniart |
| 类型 | LoKR(低秩Kronecker乘积) |
| 格式 | .safetensors(兼容LoRA) |
| Rank/Alpha | 64/32 |
| 基础模型 | Anima |
| 触发词 | @aniart |
| 推荐权重 | 0.8 - 1.2 |
快速开始
安装方法
- 下载
anima-aniart.safetensors文件 - 将其放置在你的comfyUI模型目录中(通常为
models/loras/) - 使用LoRA加载器(仅模型)加载,使用触发词
@aniart触发
基本用法
@aniart, 1girl, portrait, watercolor style, soft lighting, detailed face
推荐参数
- 权重/强度:从
0.8开始,最高可增加到1.2以获得更强效果
最佳使用场景
本LoKR最适合用于:
- 肖像和特写 - 以水彩效果美丽地捕捉面部细节
- 半身像 - 风格与构图的最佳平衡
示例提示词
肖像:
@aniart,masterpiece, best quality, very aesthetic, 1girl, close-up portrait, flowing hair, watercolor splash background, soft light, masterpiece, best quality
半身像:
@aniart,masterpiece, best quality, very aesthetic,1girl, half body, sitting, holding flowers, watercolor style, dreamy atmosphere, detailed eyes
复杂提示词:
@aniart, masterpiece, best quality, very aesthetic,
best quality, score_7, sensitive, 1girl, solo, komiya kaho, idolmaster shiny colors, red hair, ahoge, side ponytail, hair flaps, crossed bangs, red eyes, small breasts, magical girl, armored dress, (fingerless gloves:2), cloak, floating cape, bracer, breastplate, determined, raised fist, clenched hand, looking at viewer, magic circle, magic circle background, glowing, feathers, gleam, upper body
A young magical girl with red hair in a side ponytail and an ahoge stands heroically, her cloak billowing behind her as she raises a clenched fist toward the viewer with a determined expression. Her armored dress and breastplate gleam under the glow of a magic circle background, while feathers drift through the air around her. The upper-body composition frames her confident pose and sharp gaze, capturing the essence of a great hero.
局限性与建议
请注意以下几点:
- 全身构图:本LoKR会降低模型渲染全身图像时的细节表现
- 复杂背景:本LoKR会降低模型渲染复杂背景时的细节表现
- 高权重时的强风格:权重超过1.2可能会过度扭曲原始构图
获得最佳效果的技巧
- 从较低权重开始(0.6-0.8),逐渐增加以找到最佳平衡点
- 使用质量提示词:包含"masterpiece, best quality, detailed"等词语以保持图像质量
- Downloads last month
- 6
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for SAkizuki/Anima-Aniart
Base model
nvidia/Cosmos-Predict2-2B-Text2Image Finetuned
circlestone-labs/Anima

