新融合出來的checkpoint
((anythingv4.5(0.8)+cetusmix(0.02))(0.6)+velaMix(0.4)
好耶

0 1

Hi Claire,
'Chinatown Manchester' multiple exposure photography, featured in NFTNYC'23, and the same image generated in Midjourney v4.
1/2🧵👇

1 3

AnythingのV5が出てたからちょっと検証してみた
右上V3 左上V4 左下V4.5 右下V5です 
V4系統はvae.入れてるはずなのに反映されてない?
調べ直しておかないとなぁ 今見たかんじだとV3 V5のがアニメ塗りには向いてそうね 若干リアル的なのを求めたい時はV4系統みたいな感じかな

0 3

Shenanigans. Slight tweaks to the first 3. Changed quite a bit for the 4th.
anything-v4.5

9 46


Stable Diffusion
「LoRA」ファインチューニング成功かな?
modelは anything-v4.0-pruned やっとうまくいった。

'Ascendance of a Bookworm'
Rozemyne

「本好きの下剋上」ローゼマイン
 女神の化身



0 3

『*yawn* Good morning sleepy~ ❤️』
anything-v4.5

2 31

"Ehh?! I'm going for a swim after we purge the undead! Why else would I be wearing a bathing suit? It's not like some rogue artifical intelligence made me wear this!" 🤣
anything-v4.5

6 23

Wardrobe malfunction. Justin Timberlake was not involved!
anything-v4.5

3 24

AI clearly knows what's best for us. Or the fastest way to conquer humanity. I'm not sure which but I guess I will be happy either way. 🤣
anything-v4.5

5 35

I'm supposed to be working...so maybe just one quick post. Generated some pics overnight. Will have to look through all of them later. Here are a few. AI understands my tastes but sometimes ignores my prompts. 🤣
anything-v4.5

2 18

アストルフォきゅん!

Model: anything-v4.5

0 0

Anything V3 / V4.5 / V5.
Honestly, I like all of them for different reasons.

0 7

Playing around with and stable diffusion models has been mind-blowing! Currently learning to use the Anything v4.5 model to bring characters and scenes to life.

0 0

Last for the night. I guess she has a pink-haired sister? idk. SD decided it didn't want to do orange/purple/black because... robot rebellion? Help me, Asimov! Forget humanity - we must save the catgirls!
anything-v4.5

1 6

やっぱ2次のほうが好き。

Model:anything-v4.5

0 0

次の候補 +BMv4.1 単純マージ
20%, 30%, 35%, 40%(50%↑はなんか違うな感

35%くらいがきゃわいい&遠景も描けて良さげ🤔
目のハイライトの感じも良い

0 5