Built my own model in Dreambooth of my existing art to see if Stable Diffusion could closely reproduce the same images. Not quite, but it could be improved.

0 4

One of the interesting thing I discovered about is that it's absolutely possible to train for multiple subjects with good consistency. What's more, it's often possible to refer to individual subjects by minor prompt adjustment.
A 🧵:

12 77

Dreambooth James Webb Space Telescope Deep Space model by

model: https://t.co/PxCXK7BZt8

20 115



DreamBooth.
四糸乃.

新しくAnything-V3.0modelを使用しました.
새롭게 Anything-V3.0 model을 사용했습니다.

0 14

これは「なんか寒そうなイラストが見たいな…(ツム家の作業スペースは南向きで日中はまだ暑いくらい)」のお気持ちをAIに代弁してもらいました。



0 3

たーだーのくん♪ → 誰お前(頭のアスタリスクも二つになる衝撃)

1 3