This lora was trained on the Sakura No Tomoru Hi E style using 52 images. The source images had a lot of random text in them, and these text will occasionally appear if there is too much empty space in the background. Placing the tag “text” in the negative or arranging the image to fill up the space with something else should prevent this.

Version 1.0 was trained on Anima V1.0.

To get the results you see in the images above, these are the settings I used. (https://civitai.red/articles/11427/my-current-sdxl-settings)

Visit My Site For Models I Can't Post On CivitAI: https://qtgurl.com/