<p>TinAnima_v1D_INT8_4steps (Anime-style Portraits)</p><p>(add bf16 version model)</p><p>This model can generate large-scale images directly, eliminating the need for an upscaler. It offers superior quality and faster generation speeds compared to SD1.5.</p><p>Tested on an NVIDIA RTX 3060 (12GB VRAM) using WebUI Forge Neo: generating a 1024x1536 image with the INT8 model takes approximately 3.6 seconds at 4 steps.</p><p>Sampler: Euler (ER SDE)</p><p>Steps: 4 or more</p><p>CFG Scale: 1 or more</p><p>Image Size: 1024 x 1536</p><p>TinAnima_v1D_INT8_4steps 二次元人像</p><p>(增加 bf16 版本模型)</p><p>模型可以直接生成大图，可以不用放大模型，生成质量和速度都比SD1.5好和快</p><p><span style="color:rgba(255, 255, 255, 0.8)">用3060 12GB显卡测试，在WebUI Forge Neo 里int8模型生成1024x1536，4步生成用3.6秒左右</span></p><p>采样方法Sampler：Euler (ER SDE)</p><p>迭代步数Steps：4 或更多</p><p>提示词引导系数CFG：1 或更多</p><p>图片尺寸：1024 x 1536</p><p>正面提示词：可以参考图片里的参数,一般类似 Z-Image 自然语这英文的或SD1.5的单个英文字再用,分开</p><p>负面提示词用：worst quality, low quality, score_1, score_2, score_3, artist name (可不用)</p><p>提示词可以参数返图里的参数</p><p>StableDiffusion 新Anima二次元動漫人像模型分享,產生1024x1536幾秒一張圖</p><div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/tkwr_xrdN2Y" start="0"></iframe></div><p>適用于sd-webui-forge-neo的扩展TIN_batch_txt_prompt(使文件夹里的txt批量生成不同的图片)</p><div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/i9kC0zFhnac" start="0"></iframe></div><p>在SD WebUI Forge Neo 里使用</p><div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/njjc2ruwyBo" start="0"></iframe></div><p>用Anima-TrainFlow训练Anima Lora模型</p><div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/tSQILxG1UOY" start="0"></iframe></div><p>简单分享使用Anima-TrainFlow训练器来训练之前的SD1.5人像素材</p><div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/uIenrJpJqz0" start="0"></iframe></div><p><span style="color:rgba(255, 255, 255, 0.85)">(20261028 直接</span><span style="color:rgba(255, 255, 255, 0.8)">分享原模型下载</span> <span style="color:rgba(255, 255, 255, 0.85)">)</span></p><p><span style="color:rgba(255, 255, 255, 0.85)">(欢迎大家测试和返图)</span></p><p>mypikpak.com/s/VOcdFHxoJjvcCAewVVS40pQKo2 图集(20261028 civitai很慢,上传图片到这里)</p>

TinAnima_v1D_INT8_4steps 二次元人像

<p>TinAnima_v1D_INT8_4steps 二次元人像</p>

Files with SHA256: f421d830f650a8b7d2a83a79cb082b2b5e0e64bb85cf686eefddb548c829902a