<p><span style="font-size:22px">Trigger: Wacan (</span><span style="color:rgb(225, 60, 57);font-size:22px">MUST INCLUDE</span><span style="font-size:22px">)</span></p><p><span style="font-size:19px">触发词：Wacan    </span> <span style="color:rgb(225, 60, 57);font-size:19px">必须加上</span></p><h3 id="recommended-checkpoints:one-obsession-16noobai">Recommended Checkpoints:One obsession 16NoobAi</h3><p><span style="font-size:19px">以日系二次元半厚涂为基础，增强了柔和通透的光影与细腻的色彩过渡，使画面氛围清新治愈，线条干净克制</span></p><p>Based on Japanese-style anime semi-realistic painting, it enhances soft and translucent light/shadow and delicate color transitions, creating a fresh and healing atmosphere with clean, restrained line art.</p><p></p><h3 id="weight-recommendation:-0.8">Weight recommendation: 0.8</h3><p><span style="font-size:19px">权重推荐：0.8</span></p><p></p><h3 id="created-with-effort-likes-and-reposts-are-greatly-appreciated!">Created with effort — likes and reposts are greatly appreciated!</h3><p><span style="background-color:rgb(255, 255, 255);color:rgb(225, 60, 57);font-size:22px">制作不易求点赞 返图支持！</span></p><h3></h3><p></p>

Wacan anime style

<p>This is a new LoRA trained on images generated by fusing the two LoRAs I trained. The style hasn’t been fully replicated yet. I will upload the fused model later; however, it has a relatively high distortion rate, so I will keep working on style replication.</p><p>这是用我训练的两个LoRA融合出的图像，再训练而成的新LoRA。目前风格尚未完全复原，后续我会把融合模型版本上传，不过融合模型的畸形率偏高，所以我后面会继续尝试风格复刻。</p>

Files with SHA256: 4abc9e2f17b9534f3b8f88be654e2d9d8d0d2f2a691592199678814de1b68d5e