Trained with 69 pics
Recommended weight: 0.8-1
Base prompts: IlluEmma, orange hair, short hair, ahoge, green eyes,
Base Outfit: white pleated skirt, white collared shirt, untucked shirt, brown boots,
Description
Based on WAINSF, so if you use said checkpoint, you get a better anime-like style
FAQ
Comments (2)
very cool LORA. Would you like to share the training parameters?
Thank you, if you want to know, i used this guide for training, i just adjusted it to Illustrious
https://civitai.com/articles/6438/how-to-train-your-lora-sdxl-pony-with-3070rtx-8gb-of-vram-with-kohya-gui-v2415-july-2024
As for the overalls parameters, i do around 10 epochs and in each epoch i try to have between 200 and 300 steps (numbers of images X repeats = steps), and for this Lora specifically, i used waiNSF as the base checkpoint for the training.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.