这只是一个学习训练下来的结果，并不是所有方面都全面超越未加载lora的模型。改变了一下模型的构图和设计能力，同时增加了明暗对比。测试使用的权重都是1.0。正向提示词：<pre><code>masterpiece, best quality, very aesthetic,</code></pre>具体效果请参见我的示例图片。感谢大家下载并使用。如果你喜欢，欢迎关注我。<hr />This is just the result of a training run, and it does not comprehensively surpass the base model without LoRA in every aspect.It changes the model’s composition and design capabilities while also increasing light-and-dark contrast.All tests were run with a weight of 1.0.Positive prompt:<pre><code>masterpiece, best quality, very aesthetic,</code></pre>Please see my sample images for the actual results.Thank you everyone for downloading and using it. If you like it, feel free to follow me.<hr /><hr />V1：在细节方面多数情况下强于V0.8，色感更平衡。V1: In most cases, it is superior to V0.8 in terms of detail, with more balanced color perception.<hr />V0.8：明度比V0.5更高，构图动感合理优化，更明显的人物突出。 V0.8: Higher brightness than V0.5, a more dynamic and properly optimized composition, and more prominent character focus.<hr />V0.5：明暗关系更明显，一点动作和细节优化。 V0.5: More pronounced light-and-shadow contrast, with slight refinements to pose/action and details.

Aesthetic composition | 美学构图

V1多数情况下，比V0.8更强的细节，色调平衡性更好。其次V1绘制的眼睛画得更好看。In most cases, V1 has stronger detail and better tonal balance than V0.8. Secondly, the eyes rendered by V1 look better.

Files with SHA256: e8e2802293bececdd4ed7098aec33c656e9d883995d89c1502833e77986d96a8