If you like the LoRa, please like it and post your image result below... :D
I only train models for characters that I or friends of mine are interested in, so don't ask if I could train character XYZ...
A blast from the past...
To be honest it was awful to get original training data, not only was the source just in SD quality, but most of the time the character wasn't really drawn consistently, so there maybe a bit of a consistency problem in the results :/ ... this time I didn't use AI up scaling as the result was simply worse than doing normal up scaling...
Simply copy the trigger words. The base trigger word is alice mitchell. Then, add the only outfit option that stuck or use your own outfit. ^^ In case of Illustrious models I suggest putting "blurry, blur" into the negative prompt and "sharp" in the normal prompt to reduce the blurriness
A improved version which should be a bit less blurry, and looking at viewer shouldn't suck anymore :) beside that overall a bit more stable, randomly getting a blue hat or a random handbag should occur less... I will try to scale these images up a bit better as they still tend to get blurry :<
1. Hak untuk model yang diposting ulang adalah milik pembuat aslinya.
2. Pencipta asli yang ingin mengklaim model harap hubungi staf SeaArt AI melalui saluran resmi. Klik untuk mengklaim
