More flexible training and new ?????? dataset


V2 update:
New ?????? dataset and a more flexible default outfit.
nobikini, twin braids to ger her bikini;
nobeta, witch hat, dress, cape, white gloves, thighhighs for her default outfit.
The ?????? dataset only have 29 pics, so its not consistent or amazing.
===========================================================
I made a Ko-fi account, everything I make will still be free, but consider donating if you like my loras :D
Lora tested at 1 strength.
The trigger prompt is nobeta but it may be necessary to add tags such witch hat, dress, cape to the prompt to get her full outfit. Check the samples for examples. I focused the trigger word only to make her eyes and hair so she can be dressed in other outfits.
I have archived some lower epochs, let me know if you think the LoRA is too strong.
Enjoy! :D
More flexible training and new ?????? dataset
1. 轉载模型僅供學習與交流分享,其版權及最終解释權归原作者。
2. 模型原作者如需認領模型,請通過官方渠道联系海藝AI工作人員進行認證。我們致力於保護每一位創作者的權益。 點擊去認領
