<h2 id="trush-v10">Trush-DTL[Detail Tweaker LoRA] V1.0</h2>【模型细节调节LoRa︱模型手术刀】如果没有意外，这应该是最早使用差异炼丹法的SD LoRA<h3 id="or-specification">说明 | specification</h3><ul><li>This model uses a special training method,It does not contain any style or content采用差异炼丹训练法制作，本身不含有任何风格和内容</li><li>This Lora can be used to adjust the correction model.本lora可以调节模型细节</li><li>According to their own model to use positive and negative weight for adjustment, the actual test results are good根据自己的模型使用正负权重用于调节，实际测试下来效果不错</li><li>This LoRa can adjust some of the features of the model as accurately as a scalpel. To achieve the desired effect without changing the basic style of the model.这个LoRa可以像手术刀一样精准的调节模型的部分特征。使其在不改变模型基本画风的情况下达到想要的效果。</li><li>This has a very good effect on the detail explosion model that comes out of a few QQ groups.这对于少数QQ群里传出来的细节爆炸模型，有着相当好的效果。</li><li>Whether to choose A or B needs to be seen according to the specific model, some use A better effect, some use B better effect选取A还是B需要根据具体的模型来看，有的用A效果更好，有的用B效果更好。</li></ul><h2 id="heading-118">例图 | Example diagram</h2><ul><li>The example is to test the use effect (when the weight is set to -1). It is not necessary to set the weight that high in actual use例图为测试使用效果（权重开到-1的情况），实际使用没有必要开那么大权重<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/bcad77da-97b5-44d5-94cf-805b3ec21b47/width=525/bcad77da-97b5-44d5-94cf-805b3ec21b47.jpeg" /></li></ul><h2 id="or-use">使用 | Use:</h2><ul><li>Adjust the model by selecting positive and negative weights and values according to your own model situation根据自己模型情况选择正负权重以及数值，即可完成对模型的调整</li><li>This lora can be used with any prompt word for any model.For example, you can use it with <a target="_blank" rel="ugc" href="https://civarchive.com/models/28205/rainbow-lines-style">Rainbow lines style LoRa</a> to correct details此lora对于任何模型任何提示词都可以使用，例如你可以和<a target="_blank" rel="ugc" href="https://civarchive.com/models/28205/rainbow-lines-style">彩线法LoRa</a>一起使用达到如图所示的效果<img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/5db88868-eef0-43ef-d7a2-56b070dd4e00/width=960/00028-374295496.jpeg" alt="00028-374295496.png" /></li><li>This model has a powerful adjustment ability, and the specific effect is shown in the figure（The picture is an example of extreme cases）本模型具有强大的调节能力，具体效果如图(图片为极端情况)<img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/8565841a-1730-456f-f3a4-72afec071e00/width=525/8565841a-1730-456f-f3a4-72afec071e00.jpeg" /></li><li>Note that this model is careful to turn on too high a weight, otherwise the output graph will collapse completely注意，本模型谨慎开启过高的权重，否则会产生过多的细节造成输出图完全崩坏的效果（权重过高），或者使画面丢失细节（负数权重过低）<img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/65f9ff82-4978-47c2-5fc9-9d060de34c00/width=525/65f9ff82-4978-47c2-5fc9-9d060de34c00.jpeg" /><img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/1218d0b1-0974-424f-20af-a272cf015900/width=525/1218d0b1-0974-424f-20af-a272cf015900.jpeg" /></li><li>May not be compatible with partial LorAs, such as txmfkr可能不兼容部分lora，例如txmfkr等</li><li>【PS:This model can not only be used to adjust the details, but also affect the effect of the prompt.】【PS:这东西不只是可以用来调节细节，也可以影响你的提示词的效果，但是具体情况还需要自己测试（不同模型不同tag影响程度都不一样）】</li></ul><h3 id="or-limitations">局限性 | Limitations</h3><ul><li>Not compatible with live action models (not tested, training set has no map)不兼容真人模型（没测过，训练集也没对应的图）</li><li>There may be failures or crashes that are not covered by the training set可能有训练集未涉及的存在导致失效或者崩溃</li><li>As for the merged model with excessive partial fitting, it will collapse once used and cannot be used at all对于部分拟合过度的乱融的模型一用就崩，根本无法使用</li><li>The upper limit is too low(Use some of the more useful models to know, generally is not touched)上下限过低（用过一些比较好用的模型就知道了，一般是碰不到的）</li><li>This model is not suitable for fixed proportion ckpt, only suitable for external adjustment本模型不适合固定比例容入大模型，只适合外置调节使用</li></ul><h3 id="or-special-thanks">特别鸣谢 | Special Thanks</h3><ul><li><a target="_blank" rel="ugc" href="https://github.com/bmaltais/kohya_ss">GitHub - bmaltais/kohya_ss</a>：训练UI，非超好用！</li><li><a target="_blank" rel="ugc" href="https://github.com/AUTOMATIC1111/">Automatic1111</a>: SD-WebUI，帮助处理训练集，当然很大一部分图是我自己用PS裁的</li><li>Danbooru：打标器，帮助标注训练集图片，当然很大一部分图是我自己标的</li><li>秋叶的甜品店（频道）：让我发现了这个LoRa的训练方法，以及促成了这个LoRa的制作</li></ul><h3 id="3">祝大家玩得开心＜3</h3><ul><li>你可以随意将本模型合并到其他地方，但如果你共享该合并模型，别忘了标注一下We allow you to merge with another model, but if you share that merge model, don't forget to add me to the credits.</li></ul>

Files with SHA256: 3a1107d3cb3bcddf403c25b754098c1ba70e8edd8c040395a126e977ba61d14b