This LORA model is trained on delibrate v1.1. However, tested a little bad when weighted 0.65~1.0 on my usual choice like Ares Mix or Delibrate itself.Works much better when weighted 1.0 (and be careful about the unique negative prompts <s>--</s>). Feel free to give it more try~You can refer to the PNG info of each sample image for prompts. Welcome to share your works and comment^_^~~Hires.fix on is suggested as always (e.g. 448*640--1.6x--&gt;716*1024, Hires steps 18,Denoising strength 0.45, latent.)~~Be kind to these lovely.<h3>Legal &amp; Risk</h3>It is prohibited to use this model for commercial purposes and any scenarios of illegal acts and purposes.~~Feb 9: Tested good as well on <a target="_blank" rel="ugc" href="https://huggingface.co/nuigurumi/basil_mix">BasilMix</a> (---) when weighted around 0.8. (Unsolved blur and malformed hands)

Chinese Actress Li Qin 

Trained with new method. Weighted 0.6.

Files with SHA256: 452a4f3a3de607962f9d8f2d533ada65ef68d4f1c36251606e629e99dfbdfc42