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.
——

2 66

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

0 3

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.

1 4

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

6 24

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.

10 83

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.

3 24

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!

0 2

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 1.5
❤️‍🔥

0 2

been having fun mergin' diffusion models, textual inversions, and LoRAs
(more results/examples in🧵below)

2 10

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)。

0 1

Stable Diffusion standard / signature look: trained a lot with KPOP models. Ulzzang textual embed.

0 2

Our new gorgeous icon by ! And or textual icon by 痛み(XG_Itami). 🥹🫶

9 61

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。

0 0

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。

0 0

old model, with textual inversion, giving distorted arts, but still aesthetic

0 0

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。

1 1

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…
🎨

0 1

I really don't know what I've done but my latest textual inversion model gave me these

0 10


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💦

4 17