//=time() ?>
Homosexual activity in the anime I got recommend turns out to in fact be 100% textual and not copium subtext. I’m afraid I won’t survive.
——
#hamefura #mynextlifeasavillainess #katarinaclaes
Proportions break depending on the model you're using. Personally i think the faceting in the original map gives interesting textual details. Using "Hires, fix" would smooth the faceting, reducing the denoise strength help keep the original textual details
Esto pasó así textualmente. Ahorita ya creció algo y me acostumbré un poco pero el día que pasó estaba apunto de llorar. Igual me seguiré dibujando con el pelo largo porque no me lo volveré a cortar en un largo rato y volverá a estar así jeje.
#comics #viñeta #digitalart
I LOVE BERNIE! She's incredibly important to me and massively helped me contextualize my own anxiety in a way that I have finally started to learn to tackle. It might have taken funny purple anime girl, but I'm a better person because of her. (also, fanfic links in the reply) https://t.co/Rjq4ByrsnG
Birthday gift for Wolf04 on FA
To contextualize the meme, he doesn't truly have that big of a foot fetish. It's just a big meme that we tease him about in a group chat I'm in.
Tired of being asked to contextualize my works with some meaning here it goes:
- i do it because they are me, while i’m painting i’m thinking about my life, nevertheless i’m evolving with it, and they are evolving with me.
@joynxyz I'm re-contextualizing classic Indian paintings to reflect modern elements and palettes.
The original artist revolutionized Indian art by infusing western techniques into his style.
And I'm merely adding a cultural bookmark to his vision!
Some Images built with (model) Dreamshaper 3.1 (LORA) GlitchStickV15-256, WlopStyle, johnsonDesuV15-320, CrazyExpressions (Negative Textual Inversion) DangerHawk & DangerGoose
using a single pose from Jo Jo's Bizarre Pose Helper on ControlNET OpenPose for #stablediffusion 1.5
❤️🔥
been having fun mergin' diffusion models, textual inversions, and LoRAs
(more results/examples in🧵below)
untitledを少し触ってみたけど、何もしないとデッサン人形みたいな体の塗りに。
Textual Inversionはnegativeだけにしないと添付のような画風で固定されてしまう(こういうのが描きたい人には良い)。
動きのある下半身が不安定(テストするときは、いつもjumpingで様子を見る)。
Stable Diffusion standard / signature look: trained a lot with KPOP models. Ulzzang textual embed.
左上…縦長?
右上…例の超長いnegativeを入れたら逆効果
左下…左上にanimescreencapのTIを適用
右下…右上に同TIを適用
Textual Inversionすごいな。
素のモデルをテストしたいんだけど、なかなか思い通りの絵が出せません。
#PicassoDiffusion
思えば左上が一番マシだった(いつものプロンプト。設定ミスってジャギったものより彩度が落ちたため、沼に)。
右上がTextual Inversionあり。
左下がnegative無し。
右下がモデルカードのサンプルコードの長文negative入りで成功したもの。
何故か縦長になりがち。
#PicassoDiffusion
#stablediffusion #SD15 old model, with textual inversion, giving distorted arts, but still aesthetic #colorful
illuminatidiffusionについてたTextual inversionお試し(positive、左上が無し、残り3種類)。
何か変わるけど、どう変わるって表現しづらいですなー。
@The_BadDM Contextually, her adventuring companion regularly says that to her when she’s wandered off to chat someone up, staring at someone, or just making a blatantly thirsty remarks. So now she just does it more on purpose.
He still won’t bang her…
🎨 @ZomgDae
I really don't know what I've done but my latest textual inversion model gave me these
#stablediffusion #aiart #aiartcommunity
#ブルーオース
AI絵での学習中に明かされたケルンちゃんの秘密⁉️
AIの追加学習(Textual Inversion)は斬新な発想を見せてくれることがあります
キャップの王冠マークと目の輝きに着目したらしくバーサーク状態のケルンちゃんに…
⚠️一度こうなると簡単には誤解が解けません💦