<h2 id="introduction">Introduction</h2>This is my first LoRA. <a target="_blank" rel="ugc" href="https://virtualyoutuber.fandom.com/wiki/Amakami_Konomi">Amakami Konomi</a> is a Japanese female indie VTuber, illustrator, and Live2D modeler. I find her illustrations really nice, so I decided to train this.<h2 id="usage">Usage</h2>Recommended LoRA strength: 1.0Trigger word:<pre><code>amakami konomi</code></pre>Default outfit:<pre><code>blue dress, brown capelet, white frilled blouse, brown neck bow, belt, brown striped thighhighs, brown lace-up boots</code></pre><h2 id="training-details">Training Details</h2><ul><li>Base model: Anima Base 1.0</li><li>Training tool: <code>sorryhyun/anima_lora v1.17.1</code></li><li>Dataset: 7 manually captioned images</li><li>LoRA rank / alpha: 32 / 128</li><li>Learning rate: <code>2e-5</code></li><li>Optimizer: AdamW</li><li>Learning rate schedule: Cosine decay with 5% warmup</li><li>Training duration: 120 epochs / 840 steps</li><li>Precision: BF16</li><li>Resolution: Aspect-ratio buckets, approximately 1 megapixel</li><li>GPU: NVIDIA GeForce RTX 3090</li></ul><h2>简介</h2>这是我的第一个 Lora。甘狼このみ 是一个日本的女性独立 VTuber/插画家/Live2D 建模师。我觉得她的插画很可爱，所以我训练了这个 Lora。<h2>使用</h2>激活权重：1.0激活词：<pre><code>amakami konomi</code></pre>常服：<pre><code>blue dress, brown capelet, white frilled blouse, brown neck bow, belt, brown striped thighhighs, brown lace-up boots</code></pre><h2>训练参数</h2><ul><li>基础模型：Anima Base 1.0</li><li>训练工具：<code>sorryhyun/anima_lora v1.17.1</code></li><li>训练数据：7 张手工标注图片</li><li>LoRA rank / alpha：32 / 128</li><li>学习率：<code>2e-5</code></li><li>优化器：AdamW</li><li>学习率调度：余弦衰减，预热比例 5%</li><li>训练量：120 epochs / 840 steps</li><li>训练精度：BF16</li><li>训练分辨率：按长宽比分桶，约 1 M 像素</li><li>训练显卡：NVIDIA GeForce RTX 3090</li></ul>

Files with SHA256: e60be6c43863d084fc6999126611135562a3d942cca3243c1ff4c81e0a5edcec