Anon’s Dream Scene is a 2D anime-style scenery LoRA model. It is designed to enhance the visual aesthetics of anime scene while reducing their realistic look. Anon’s Dream Scene 是一个 2D 动漫风格的场景 LoRA 模型，用于增强 anima 模型的动漫场景视觉美感，并降低其写实倾向。Trigger word: <code>anon_scene</code> 触发词：<code>anon_scene</code>You can <a target="_blank" rel="ugc" href="https://sansui233.github.io/static/anon-scene-lora-tags/">find the full tag frequencies on my blog</a>. This LoRA is designed to be triggered easily with common scenery-related tags. For examples:<ul><li>outdoors: <code>anon_scene, scenery, outdoors, nature</code></li><li>indoors: <code>anon_scene, indoor</code></li></ul>你可以在我的博客中查看完整的标签频率统计。该 LoRA 设计为可通过常见的场景类标签轻松触发，例如：<ul><li>室外：anon_scene, scenery, outdoors, nature（场景、室外、自然）</li><li>室内：anon_scene, indoor（室内）</li></ul>The style is intentionally a little overfitted. You can lower the LoRA weight (&lt;0.7) to get a better balance when using it with other LoRAs.该风格有意进行了轻度过拟合处理。在与其他 LoRA 叠加使用时，可以将权重降低到 0.7 以下以获得更好的平衡效果。<h2 id="style-description">Style Description 风格描述</h2>Indoors: Bright lighting, clear shapes, and sharp line details. 室内：明亮的光照、清晰的形体结构、较锐利的线条细节。Outdoors: Bright anime-style colors. 室外：明亮的日系动漫风格色彩。The model was trained on natural photos, animation screenshots, scenery illustrations, and 3D models with cartoon-style rendering. 该模型的训练数据包含自然照片、动画截图、场景插画，以及卡通渲染风格的 3D 模型。

Anon's dream scene - Anime background enhancement

This version is strongly anime style oriented.

anon-scene_v1.0.safetensors

Anon's dream scene - Anime scenery enhancement

v1.0

v1.0-extra

Trained on v1.0. Denoised to address overfitting in street scenes and similar scenarios. This version has a stronger anime style, so please use it with caution.

Files with SHA256: 748a565b0b737eda42d5d3efb1436929bb4c3665774ae50e96364f52d4493348