//=time() ?>
1. orange+onimai model-AC
2. orange+ExtratLora(onimai model-AC)
3. orange+trained onimai lora
I believe not having the weight difference of the residual blocks is very detrimental
This is also why I always train entire model and merge them
#stablediffusion #lora
Another Squigly image from the same version of the Lora as the previous image. This image is insanely busy but I thought it looked cool enough to share. #AIart #stablediffusion #Skullgirls
https://t.co/6NyWWM1MuH
You're not dreaming—I'm a Star Guardian now.
#LeagueOfLegendsFanArt #AIArtwork #AIart #lora #LeagueOfLegends #Jinx #StarGuardian #starguardianJinx #grapefruitmix
プリズマイリヤのLoRA学習ファイルをダウンロードしたので、ツヴァイフォームのイリヤちゃんを生成してみました。
#AIart #プリズマイリヤ #FateGO #FGO #FateGrandOrder #イリヤスフィール・フォン・アインツベルン #LoRA
Replicant-V1.0+LoRA
t2i(Hires. fix)
#AIart #StableDiffusion #WaifuDiffusion #Dreambooth
自分の過去絵を食わせてLORA作ってみたんだけど枚数が少ないからかなんかよくわからんな…重み増すと主線が強くなってちょっとそれっぽくなるけど人体狂っちゃうな
#AIイラスト
#stablediffusion #GenshinImpact Another Lora model of Sangonomiya Kokomiby me: https://t.co/gCgdi83Srw trained by the 5+5 method. So 5+5 is half validated.
#stablediffusion #GenshinImpact my first trained well-working LoRa model: https://t.co/DuRrpKPxCX It was the first time to find the 5+5 LoRa training method on this model. Two costumes of Ningguang were merged into one Lora.
And most importantly, the face is so official!