This is an embedding of female character that I've generated or merged by embedding merging extensions thru SD based on 3rd-party asian doll likeness loras.
Since this refered person doesn't exist in the real world, any resemblance to anyone in the real world is unintentional and coincidental.
Description
Train with base model v1-5.ckpt，200 steps , 600 epochs.
歡迎各位老色批，不發（看）露點圖你上什麼CIVITAI對吧。
CIVITAIへようこそ痴漢の皆さんヌード写真を見に来たのですね。
FAQ
Comments (4)
curious.. how did you train this with less steps than epochs?
Because I set the [Batch size] number to 3, that means there're 3 packs of photos(datasheet) being trained in each cycle. See my training settings thanks to JernauGurgeh. BTW, I train her uesing only 1 photo, yes, 1 PHOTO ONLY. Is that amazing?
@AceNeo i see. havnt tried that. does that only work for embeddings? how does SD study the different angles with 1 photo though?
@ryk_civit Not sure of other type of trainings, since I only do T.I.(embedding) training. I think SD detect parameter from the face dimension, after learning process, it does estimation through algorithm, then fill up the edge factors of face, final complete whole model.