大家好，我是TuTuCSF，我定期在<a target="_blank" rel="ugc" href="https://huggingface.co/TuTuCSF/stable-diffusion-lora-Lolita">huggingface</a>发布LoRA和模型.Hello everyone, I am TuTuCSF and I regularly publish LoRA and models on <a target="_blank" rel="ugc" href="https://huggingface.co/TuTuCSF/stable-diffusion-lora-Lolita">huggingface</a>.这款洛拉模型是为了追求那些美丽的洛丽塔连衣裙而设计的，动漫和现实主义都能产生良好的效果。在0.8和1.0之间效果最好This Lora model is created to pursue those beautiful Lolita dresses, and both anime and realism can produce good effects.Works best between 0.8 &amp; 1.0

【北境夜雪】Dress No.3 Blue Dress

我发现1.3版本有概率会出现照片残缺的缺陷，重新调整了训练集，解决了这个问题，训练了两个版本，这个是a版本，b版本可以在<a target="_blank" rel="ugc" href="https://huggingface.co/TuTuCSF/stable-diffusion-lora-Lolita">TuTuCSF/stable-diffusion-lora-Lolita · Hugging Face</a>获取I found that version 1.3 has a probability of having photo defects. I adjusted the training set to address this issue and trained two versions. This is version a, and version b can be obtained on <a target="_blank" rel="ugc" href="https://huggingface.co/TuTuCSF/stable-diffusion-lora-Lolita">TuTuCSF/stable-diffusion-lora-Lolita · Hugging Face</a>

Files with SHA256: 8eb02b563e3aa534fa0b1ad9205abde0ab7103d1a8dd52062a5fe61b751c38f9