Trained from "The Super Mario Bros. Movie"
If you want to get a close result, a clear image, let the checkpoint & Lora do its work. Extra tags like realistic, photoreal, etc. will only damage the likeness of the character. Less you tags for better results.
If you can't get close results, you can use the word "mpeach" to enhance the Lora.
There is a huge problem in v1 tagging, Her keep getting the walking pose, and some time stuck with the stick, or spear weapon + the lobby background. I hope the new version fixed those issue, And I adding more image, for better fit.
Little tune to the training for a better fit for the " GoodFIt v2".
Testing all my lora on my own checkpoint "GoodFit" other 3d checkpoints might work but maybe showing different results,
Trained her with 2 extra outfits, you can use the words " bodysuit" and "wedding" to trigger it. Those two words are not meant to be working together.
Give it a try, hopes you like it :)
share some post and join the discord :D
If there is any trigger word or training problem let me know, I will try my best to fix it (might takes some time) :3
描述:
训练词语:
名称: mpeachv3.safetensors
大小 (KB): 295027
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
