this is my first working embedding. it can work with many models and give great results. have fun!
Description
FAQ
Comments (11)
Do you mind sharing your training settings? Training embedding gives me headache.
sure.
those are my training settings
https://imageshack.com/i/pmII4bqYp
https://imageshack.com/i/pmjsGzuCp
i trained on this model
https://huggingface.co/a1079602570/animefull-final-pruned/blob/main/novelailatest-pruned.ckpt
and i chaneged the style file words text file to the one from the list from this post
https://www.reddit.com/r/StableDiffusion/comments/xvzlxs/some_tips_for_textual_inversion_with_human_subject/
have fun!
This is the result of 5000 steps?
What is your init token? Girl or woman?
How many images?
i didn't complete the 5000 stpes beacaue i was already stasfied with the results. and i used woman
10 images?
i used like 45 images
Thank you. I'll ask you again later if I get confused.
what model did you use as base, and what are the prompts of those photos
i used this model:https://huggingface.co/a1079602570/animefull-final-pruned/blob/main/novelailatest-pruned.ckpt
you can check the prompts of the photos if you drag them to the PNG Info tab
negative prompt is crazy long
he just used: (EMB_sksmakima2-2075),solo,1girl(samdoesart1:0.5)
yep, embeds are a powerful tool
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.









