タイプ
LORA
投稿日時
2023-05-06
基本モデル
SD 1.5
トリガーワード
amee
コピー






This is my first LoRa, featuring a famous singer from Vietnam, Amee. I only trained this LoRa with just 19 pics of her so ofc it won't be perfect... yet.
And because of the first time i tried to train something so please leave a comment so i can improve in the future.
I haven't found out the perfect parameters for this, so experiment with anything you want.
Beside that you could use the parameters that i used in most of the renders:
Steps: 30,
Sampler: DDIM,
CFG scale: 7,
Size: 512x768,
Model: lofi_V2pre,
Denoising strength: 0.34,
Clip skip: 2, ENSD: 31337,
Hires upscale: 2,
Hires upscaler: Latent,
Eta DDIM: 0.67
1.転載モデルは学習・共有目的のみで使用し、著作権は原作者に帰属します
2.モデルの認証は公式チャンネルでご連絡ください。クリエイターの権利保護に努めています クリックして認証
