LEOSAM's 兔狲/Pallas's cat/manul/マヌルネコ LoRA v2.0

SD 1.5 Standard LORA
Download

LEOSAM's 兔狲/Pallas's cat/manul/マヌルネコ LoRA v2.0

2.0版本相比1.0版本的变化：

  1. 更新了训练图集，300张训练图中有约70张被替换。原图集中的模糊照片被替换。增加了兔狲幼崽照片的比例。

  2. 训练图的像素从512*512提升至640*640。

  3. 新的LORA模型提升network_dim值至160。

  4. 采用了新的自适应优化器来训练模型。

经过我的初步测试，2.0版本相比1.0版本，提升了生成逼真兔狲图片的成功概率

🖥️Welcome to try out the open-source GPT4V-Image-Captioner, developed by my friend and me. It offers a one-click installation and comes integrated with multiple features including image pre-compression, image tagging, and tag statistics. Recently, we also launched the webui plugin version of this tool, everyone is welcome to use it!

🌍欢迎加入QQ群"兔狲·AIGC梦工北厂"，群号 ：780132897 ；"兔狲·AIGC梦工南厂"，群号 ：835297318（入群答案：兔狲）。Telegram群聊“兔狲的SDXL百老汇”，链接：https://t.me/+KkflmfLTAdwzMzI1

这是一个用于生成兔狲（Pallas's cat）AI图片的LoRA

注意事项：

  1. 与MoonMix reality2.0或其他写实模型搭配使用，可以生成逼真的兔狲照片。

  2. 搭配其他底模，生成其他画风兔狲照片的效果也不错。

  3. 推荐的CFG范围：5~7

  4. 推荐的Lora权重为0.5~0.7

This is a LoRA for generating AI images of Pallas's cat.

Precautions:

  • When used in combination with MoonMix reality2.0 or other photorealistic models, it can generate lifelike Pallas's cat photos.

  • Combining it with other base models can also produce good results for Pallas's cat photos in different art styles.

  • Recommended CFG range: 5~7

  • Recommended LoRA weight: 0.5~0.7